author
int64 658
755k
| date
stringlengths 19
19
| timezone
int64 -46,800
43.2k
| hash
stringlengths 40
40
| message
stringlengths 5
490
| mods
list | language
stringclasses 20
values | license
stringclasses 3
values | repo
stringlengths 5
68
| original_message
stringlengths 12
491
|
|---|---|---|---|---|---|---|---|---|---|
718,767 |
19.04.2021 21:07:00
| -10,800 |
c090fc40ea60aa5ade7d83f17cf5d9cbece662fb
|
FIX Issue added check to prevent divide by zero error in "printDiscoveredStats".
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -815,7 +815,8 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nnumPointers,\nnumVtables) = self.getDiscoveredInfo()\n- self.vprint(\"Percentage of discovered executable surface area: %.1f%% (%s / %s)\" % (disc*100.0/(disc+undisc), disc, disc+undisc))\n+ percentage = disc*100.0/(disc+undisc) if disc or undisc else 0\n+ self.vprint(\"Percentage of discovered executable surface area: %.1f%% (%s / %s)\" % (percentage, disc, disc+undisc))\nself.vprint(\" Xrefs/Blocks/Funcs: (%s / %s / %s)\" % (numXrefs, numBlocks, numFuncs))\nself.vprint(\" Locs, Ops/Strings/Unicode/Nums/Ptrs/Vtables: (%s: %s / %s / %s / %s / %s / %s)\" % (numLocs, numOps, numStrings, numUnis, numNumbers, numPointers, numVtables))\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
FIX Issue #391, added check to prevent divide by zero error in "printDiscoveredStats". (#392)
|
718,765 |
03.05.2021 16:57:53
| 14,400 |
58e87801a405e8d4f56f2a435ff30386572b4dff
|
Prepare for v1.0.2 release
|
[
{
"change_type": "MODIFY",
"old_path": ".bumpversion.cfg",
"new_path": ".bumpversion.cfg",
"diff": "[bumpversion]\n-current_version = 1.0.1\n+current_version = 1.0.2\ncommit = True\ntag = True\ntag_message =\n"
},
{
"change_type": "MODIFY",
"old_path": "CHANGELOG.rst",
"new_path": "CHANGELOG.rst",
"diff": "Vivisect Changelog\n******************\n+v1.0.2 - 2021-05-02\n+===================\n+\n+Features\n+--------\n+- Refactor and update the posix impapi\n+ (`#390 <https://github.com/vivisect/vivisect/pull/390>`_)\n+\n+Fixes\n+-----\n+- Ancient visgraph bug\n+ (`#387 <https://github.com/vivisect/vivisect/pull/387>`_)\n+- Easier version engineering\n+ (`#388 <https://github.com/vivisect/vivisect/pull/388>`_)\n+- Remove Travis CI config and fully cut over to Circle CI\n+ (`#389 <https://github.com/vivisect/vivisect/pull/389>`_)\n+- Add check to prevent divide by zero in print stats\n+ (`#392 <https://github.com/vivisect/vivisect/pull/392>`_)\n+- Fix SaveToWorkspaceServer\n+ (`#393 <https://github.com/vivisect/vivisect/pull/393>`_)\n+- Intel emulator bug fixes\n+ (`#394 <https://github.com/vivisect/vivisect/pull/394>`_)\n+- Tests for intel emulator and more fixes\n+ (`#395 <https://github.com/vivisect/vivisect/pull/395>`_)\n+\n+\nv1.0.1 - 2021-04-05\n===================\n"
},
{
"change_type": "MODIFY",
"old_path": "README.md",
"new_path": "README.md",
"diff": "@@ -103,7 +103,6 @@ be loadable in python3 vivisect.\n## Build Status\n[](https://circleci.com/gh/vivisect/vivisect/tree/master)\n-[](https://travis-ci.org/vivisect/vivisect)\n## Extending Vivisect / Vdb\n"
},
{
"change_type": "MODIFY",
"old_path": "setup.py",
"new_path": "setup.py",
"diff": "from setuptools import find_packages, setup\nfrom os import path\n-VERSION = '1.0.1'\n+VERSION = '1.0.2'\ndirn = path.abspath(path.dirname(__file__))\nwith open(path.join(dirn, 'README.md'), 'r') as fd:\n"
},
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -2291,6 +2291,6 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these yourself\n-version = (1, 0, 1)\n+version = (1, 0, 2)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -3086,6 +3086,6 @@ def getVivPath(*pathents):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these directly\n-version = (1, 0, 1)\n+version = (1, 0, 2)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Prepare for v1.0.2 release (#398)
|
718,768 |
06.05.2021 09:54:12
| 21,600 |
1cd994946a3ca7d43ed7e1cb1146a27c2f499181
|
ci: check style via black and isort
see
|
[
{
"change_type": "ADD",
"old_path": null,
"new_path": ".github/workflows/style.yml",
"diff": "+name: style\n+\n+on:\n+ push:\n+ branches: [ master ]\n+ pull_request:\n+ branches: [ master ]\n+\n+jobs:\n+ code_style:\n+ runs-on: ubuntu-latest\n+ steps:\n+ - name: Checkout viv\n+ uses: actions/checkout@v2\n+ - name: Set up Python 3.8\n+ uses: actions/setup-python@v2\n+ with:\n+ python-version: 3.8\n+ - name: Install dependencies\n+ run: pip install 'isort==5.*' black\n+ - name: Lint with isort\n+ run: isort --profile black --length-sort --line-width 120 -c .\n+ - name: Lint with black\n+ run: black -l 120 --check .\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
ci: check style via black and isort (#356)
see #294
Co-authored-by: James Gross <45212823+rakuy0@users.noreply.github.com>
|
718,765 |
31.07.2021 00:09:44
| 14,400 |
6de41d2f238ec4531ae0f85b57ea9d11ac8d5f9e
|
Fix vstruct number issue
Squash a few bugs in vstruct related to negative numbers and unformatted values.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/teststruct.py",
"new_path": "vivisect/tests/teststruct.py",
"diff": "@@ -28,4 +28,7 @@ class StructTest(unittest.TestCase):\nself.assertEqual(vs.foo, 0x03020100)\nself.assertEqual(vs.bar, 0x07060504)\nself.assertEqual(vs.baz, 0x0b0a0908)\n- self.assertEqual(vs.biz, 0xfffefdfc)\n+ self.assertEqual(vs.biz, -0x010204)\n+\n+ fd.seek(0)\n+ self.assertEqual(vs.vsEmit(), fd.read())\n"
},
{
"change_type": "MODIFY",
"old_path": "vstruct/primitives.py",
"new_path": "vstruct/primitives.py",
"diff": "@@ -123,16 +123,33 @@ class v_prim(v_base):\nnum_fmts = {\n+ # big endian unsigned\n(True, 1): '>B',\n(True, 2): '>H',\n(True, 4): '>I',\n(True, 8): '>Q',\n+\n+ # little endian unsigned\n(False, 1): '<B',\n(False, 2): '<H',\n(False, 4): '<I',\n(False, 8): '<Q',\n}\n+signed_fmts = {\n+ # big endian signed\n+ (True, 1): '>b',\n+ (True, 2): '>h',\n+ (True, 4): '>i',\n+ (True, 8): '>q',\n+\n+ # little endian signed\n+ (False, 1): '<b',\n+ (False, 2): '<h',\n+ (False, 4): '<i',\n+ (False, 8): '<q',\n+}\n+\nclass v_number(v_prim):\n_vs_length = 1\n@@ -140,13 +157,13 @@ class v_number(v_prim):\ndef __init__(self, value=0, bigend=False, enum=None):\nv_prim.__init__(self)\nself._vs_bigend = bigend\n- self._vs_value = value\nself._vs_enum = enum\nself._vs_length = self.__class__._vs_length\nself._vs_fmt = num_fmts.get((bigend, self._vs_length))\n# TODO: could use envi.bits, but do we really want to dep on envi?\nself.maxval = (2 ** (8 * self._vs_length)) - 1\n+ self.vsSetValue(value)\ndef vsGetValue(self):\nreturn self._vs_value\n@@ -171,7 +188,7 @@ class v_number(v_prim):\nelse:\nr = []\nfor i in range(self._vs_length):\n- r.append(ord(fbytes[offset + i]))\n+ r.append(fbytes[offset + i])\nif not self._vs_bigend:\nr.reverse()\n@@ -328,13 +345,13 @@ class v_number(v_prim):\nclass v_snumber(v_number):\n_vs_length = 1\n- def __init__(self, value=0, bigend=False):\n- v_number.__init__(self, value=value, bigend=bigend)\n-\n- # TODO: could use envi.bits, but do we really want to dep on envi?\n+ def __init__(self, value=0, bigend=False, enum=None):\nsmaxval = (2**((8 * self._vs_length)-1)) - 1\nself.smask = smaxval + 1\n+ v_number.__init__(self, value=value, bigend=bigend, enum=enum)\n+ self._vs_fmt = signed_fmts.get((bigend, self._vs_length))\n+\ndef vsSetValue(self, value):\nvalue = value & self.maxval\nif value & self.smask:\n"
},
{
"change_type": "MODIFY",
"old_path": "vstruct/tests/testbasic.py",
"new_path": "vstruct/tests/testbasic.py",
"diff": "@@ -299,3 +299,44 @@ class VStructTest(unittest.TestCase):\ns.vsParse(bytez)\nself.assertEqual(s.vsGetValue(), \"accelerator.dll\")\nself.assertEqual(s.vsEmit(), b'a\\x00c\\x00c\\x00e\\x00l\\x00e\\x00r\\x00a\\x00t\\x00o\\x00r\\x00.\\x00d\\x00l\\x00l\\x00\\x00\\x00')\n+\n+ def test_unsigned(self):\n+ s = v_uint16(value=17)\n+ byts = s.vsEmit()\n+ self.assertEqual(b'\\x11\\x00', byts)\n+\n+ new = v_uint16()\n+ new.vsParse(byts)\n+ self.assertEqual(new.vsEmit(), s.vsEmit())\n+\n+ # test class that skips by having a real format\n+ class v_silly(v_number):\n+ _vs_builder = True\n+ _vs_length = 5\n+\n+ num = v_silly(value=1947824, bigend=True)\n+ byts = num.vsEmit()\n+ self.assertEqual(b'\\x00\\x00\\x1d\\xb8\\xb0', byts)\n+\n+ othr = v_silly(bigend=True)\n+ othr.vsParse(byts)\n+ self.assertEqual(othr, num)\n+\n+ def test_signed(self):\n+ s = v_int16(value=-1)\n+ self.assertEqual(b'\\xff\\xff', s.vsEmit())\n+\n+ s.vsSetValue(1)\n+ self.assertEqual(b'\\x01\\x00', s.vsEmit())\n+\n+ s = v_int16(value=-1, bigend=True)\n+ self.assertEqual(b'\\xff\\xff', s.vsEmit())\n+\n+ s.vsSetValue(1)\n+ self.assertEqual(b'\\x00\\x01', s.vsEmit())\n+\n+ s = v_int32(value=32)\n+ self.assertEqual(b'\\x20\\x00\\x00\\x00', s.vsEmit())\n+\n+ s = v_int32(value=-32)\n+ self.assertEqual(b'\\xe0\\xff\\xff\\xff', s.vsEmit())\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Fix vstruct number issue (#412)
Squash a few bugs in vstruct related to negative numbers and unformatted values.
|
718,768 |
10.08.2021 12:18:23
| 21,600 |
2179f686d3fcc4995c3ca503f075f890f3eeec2a
|
pe: better handle invalid export filename
if cannot be parsed, return None as documented.
decode as ASCII, as documented by Microsoft.
closes
|
[
{
"change_type": "MODIFY",
"old_path": "PE/__init__.py",
"new_path": "PE/__init__.py",
"diff": "@@ -454,7 +454,13 @@ class PE(object):\n'''\nif self.IMAGE_EXPORT_DIRECTORY is not None:\nrawname = self.readAtRva(self.IMAGE_EXPORT_DIRECTORY.Name, 32)\n- return rawname.split(b'\\x00')[0].decode('utf-8')\n+ if not rawname:\n+ return None\n+\n+ try:\n+ return rawname.partition(b'\\x00')[0].decode('ascii')\n+ except UnicodeDecodeError:\n+ return None\nreturn None\ndef getImports(self):\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
pe: better handle invalid export filename (#426)
if cannot be parsed, return None as documented.
decode as ASCII, as documented by Microsoft.
closes #425
|
718,765 |
10.08.2021 14:40:57
| 14,400 |
6453b76b9946239ef7603294d6eac2ac3f151837
|
Change amd64 symboliks callconv class declaration
Reorder Amd64 symboliks callconv declaration to avoid exceptions
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/symboliks/archs/amd64.py",
"new_path": "vivisect/symboliks/archs/amd64.py",
"diff": "@@ -212,10 +212,10 @@ class Amd64SymbolikTranslator(vsym_i386.IntelSymbolikTranslator):\nclass Amd64ArgDefSymEmu(vsym_i386.ArgDefSymEmu):\n__xlator__ = Amd64SymbolikTranslator\n-class MSx64CallSym(e_amd64.MSx64Call, vsym_callconv.SymbolikCallingConvention):\n+class MSx64CallSym(vsym_callconv.SymbolikCallingConvention, e_amd64.MSx64Call):\n__argdefemu__ = Amd64ArgDefSymEmu\n-class SysVAmd64CallSym(e_amd64.SysVAmd64Call, vsym_callconv.SymbolikCallingConvention):\n+class SysVAmd64CallSym(vsym_callconv.SymbolikCallingConvention, e_amd64.SysVAmd64Call):\n__argdefemu__ = Amd64ArgDefSymEmu\nmsx64callsym = MSx64CallSym()\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/symboliks/common.py",
"new_path": "vivisect/symboliks/common.py",
"diff": "@@ -2,7 +2,6 @@ import hashlib\nimport operator\nimport functools\nimport itertools\n-import collections\nimport vstruct\nimport envi.bits as e_bits\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/symboliks/tests/test_intel.py",
"new_path": "vivisect/symboliks/tests/test_intel.py",
"diff": "import unittest\n+import vivisect.const as v_const\n+\n+import vivisect.tools.graphutil as v_t_graph\n+\nimport vivisect.symboliks.archs.i386 as i386sym\nimport vivisect.symboliks.analysis as v_s_analysis\n@@ -25,6 +29,7 @@ class IntelSymTests(unittest.TestCase):\n@classmethod\ndef setUpClass(cls):\ncls.i386_vw = helpers.getTestWorkspace('linux', 'i386', 'vdir.llvm')\n+ #cls.gcc_vw = helpers.getTestWorkspace('linux', 'amd64', 'gcc-7')\ndef test_constraints(self):\nfva = 0x080509e0\n@@ -165,3 +170,23 @@ class IntelSymTests(unittest.TestCase):\n]\nfor ret, effects in sctx.getSymbolikOutputs(fva):\nself.assertTrue((ret, effects) in out, msg='(%s, %s) is not in the symbolik outputs' % (ret, effects))\n+\n+ def SKIPtest_symbolik_gcc_subchildren(self):\n+ '''\n+ So before intify in vivisect/symboliks/common.py, this would bomb out on int has no\n+ attribute \"kids\" because the dellocate callspace function wasn't properly marshalling\n+ argc dellocation as Const() (and instead as raw ints)\n+\n+ This is here mostly for codification purposes, since actually running this consumes\n+ more RAM than CI can handle.\n+ '''\n+ vw = self.gcc_vw\n+ fva = 0x00406ad6\n+ sctx = v_s_analysis.getSymbolikAnalysisContext(vw, consolve=True)\n+ graph = sctx.getSymbolikGraph(fva)\n+ for path in v_t_graph.getLongPath(graph):\n+ for emu, effects in sctx.getSymbolikPaths(fva, paths=[path, ], graph=graph):\n+ break\n+\n+ calls = [x for x in effects if x.efftype == v_const.EFFTYPE_CALLFUNC]\n+ self.assertEqual(len(calls), 26)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Change amd64 symboliks callconv class declaration (#413)
Reorder Amd64 symboliks callconv declaration to avoid exceptions
|
718,770 |
12.08.2021 06:48:51
| 14,400 |
ad6fbba99a6e82ddefbbf8137e7e4e450fe5fc11
|
impapi windows/arm and winkern/arm
|
[
{
"change_type": "ADD",
"old_path": null,
"new_path": "vivisect/impapi/windows/arm.py",
"diff": "+import vivisect.impapi.windows.i386 as v_w_i386\n+apitypes = dict(v_w_i386.apitypes)\n+\n+api = {\n+}\n+\n+i386_omits = set([\n+ 'ntdll.seh3_prolog',\n+ 'ntdll.seh4_prolog',\n+ 'ntdll.seh4_gs_prolog',\n+ 'ntdll.seh3_epilog',\n+ 'ntdll.seh4_epilog',\n+ 'ntdll.eh_prolog',\n+ 'ntdll.gs_prolog',\n+])\n+\n+for normname,(rtype,rname,cconv,cname,cargs) in v_w_i386.api.items():\n+ if normname in i386_omits:\n+ continue\n+\n+ api[normname] = (rtype,rname,'armcall',cname,cargs)\n+\n"
},
{
"change_type": "ADD",
"old_path": null,
"new_path": "vivisect/impapi/winkern/arm.py",
"diff": "+import vivisect.impapi.winkern.i386 as v_k_i386\n+\n+apitypes = dict(v_k_i386.apitypes)\n+\n+api = {}\n+for normname, (rtype, rname, cconv, cname, cargs) in v_k_i386.api.items():\n+ api[normname] = (rtype, rname, 'armcall', cname, cargs)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testimpapi.py",
"new_path": "vivisect/tests/testimpapi.py",
"diff": "@@ -14,9 +14,16 @@ class ImpApiTest(unittest.TestCase):\nimp = viv_impapi.getImportApi('windows','i386')\nself.assertEqual( imp.getImpApiCallConv('ntdll.RtlAllocateHeap'), 'stdcall')\n+ imp = viv_impapi.getImportApi('windows','arm')\n+ self.assertEqual( imp.getImpApiCallConv('ntdll.RtlAllocateHeap'), 'armcall')\n+\n#def test_impapi_posix(self):\n#imp = viv_impapi.getImpApi('windows','i386')\ndef test_impapi_winkern(self):\nimp = viv_impapi.getImportApi('winkern','i386')\nself.assertEqual( imp.getImpApiCallConv('ntoskrnl.ObReferenceObjectByHandle'), 'stdcall')\n+\n+ imp = viv_impapi.getImportApi('winkern','arm')\n+ self.assertEqual( imp.getImpApiCallConv('ntoskrnl.ObReferenceObjectByHandle'), 'armcall')\n+\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
impapi windows/arm and winkern/arm (#431)
|
718,768 |
20.08.2021 17:33:48
| 21,600 |
0c1ddbdc7e35183b7f15dfb17592a8e3464c22d2
|
pe: better handle invalid import name
closes
|
[
{
"change_type": "MODIFY",
"old_path": "PE/__init__.py",
"new_path": "PE/__init__.py",
"diff": "@@ -903,8 +903,13 @@ class PE(object):\nidx+=1\ncontinue\n+ try:\nfuncname = ibn.Name\n+ except UnicodeDecodeError:\n+ funcname = None\n+ logger.warning(\"pe: failed to read import name at RVA 0x%x\", ibn_rva)\n+ if funcname is not None:\nimports_list.append((save_name + arrayoff, libname, funcname))\nidx += 1\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
pe: better handle invalid import name (#441)
closes #440
|
718,765 |
22.08.2021 04:03:13
| 14,400 |
08b3c5212bbec1ca220798b0233dd7bbf0735e51
|
V104 release prep
Prep for releasing v1.0.4
|
[
{
"change_type": "MODIFY",
"old_path": ".bumpversion.cfg",
"new_path": ".bumpversion.cfg",
"diff": "[bumpversion]\n-current_version = 1.0.3\n+current_version = 1.0.4\n[bumpversion:file:setup.py]\nsearch = VERSION = '{current_version}'\n"
},
{
"change_type": "MODIFY",
"old_path": "CHANGELOG.rst",
"new_path": "CHANGELOG.rst",
"diff": "Vivisect Changelog\n******************\n+v1.0.4 - 2021-08-22\n+===================\n+\n+Features\n+--------\n+- Add structures to UI and a compressed version of the file to the meta events.\n+ (`#396 <https://github.com/vivisect/vivisect/pull/396`_)\n+- Actual documentation!\n+ (`#400 <https://github.com/vivisect/vivisect/pull/400>`_)\n+- Massive ELFPLT overhaul.\n+ (`#401 <https://github.com/vivisect/vivisect/pull/401>`_)\n+- Speed tweaks for the pointers pass and the workspace emulator.\n+ (`#402 <https://github.com/vivisect/vivisect/pull/402>`_)\n+\n+Fixes\n+-----\n+- RTD didn't like python 3.9, so go with 3.8.\n+ (`#400 <https://github.com/vivisect/vivisect/pull/400>`_)\n+- Have ud2 on amd64 halt codeflow and fix a MACH-O bug.\n+ (`#403 <https://github.com/vivisect/vivisect/pull/403>`_)\n+- Fix issues in vtrace's windows, vivisect/reports, PE/carve, and others.\n+ (`#404 <https://github.com/vivisect/vivisect/pull/404>`_)\n+- Tons of i386 emulator fixes.\n+ (`#405 <https://github.com/vivisect/vivisect/pull/405>`_)\n+- Safeguard mnemonic counting in codeblocks.py.\n+ (`#408 <https://github.com/vivisect/vivisect/pull/408>`_)\n+- Fix funcgraph issues with line highlighting.\n+ (`#409 <https://github.com/vivisect/vivisect/pull/409>`_)\n+- Fix issues in i386 decoding, a new thunk pass, new ELF relocations support, and more.\n+ (`#411 <https://github.com/vivisect/vivisect/pull/411>`_)\n+- Fix vstruct signed number issue.\n+ (`#412 <https://github.com/vivisect/vivisect/pull/412>`_)\n+- Change AMD64 symboliks class declaration to get the right dealloc method.\n+ (`#413 <https://github.com/vivisect/vivisect/pull/413>`_)\n+- Remove wintypes import for vtrace to avoid a python bug.\n+ (`#416 <https://github.com/vivisect/vivisect/pull/416>`_)\n+- Raise specific exception on invalid architecture.\n+ (`#418 <https://github.com/vivisect/vivisect/pull/418>`_)\n+- Raise specific exception on invalid section alignment.\n+ (`#420 <https://github.com/vivisect/vivisect/pull/420>`_)\n+- Raise specific exception on corrupt file.\n+ (`#422 <https://github.com/vivisect/vivisect/pull/422>`_)\n+- Better handle invalid exported filename in PE files.\n+ (`#426 <https://github.com/vivisect/vivisect/pull/426>`_)\n+- Fix struct.unpack issue and float issue on corrupt files.\n+ (`#428 <https://github.com/vivisect/vivisect/pull/428>`_)\n+- ARM impapi files.\n+ (`#431 <https://github.com/vivisect/vivisect/pull/431>`_)\n+- Fix python 3.8 compatibility issues (and add to CI) and fix platformDetach.\n+ (`#432 <https://github.com/vivisect/vivisect/pull/432>`_)\n+- Alignment and padding of PE sections.\n+ (`#436 <https://github.com/vivisect/vivisect/pull/436>`_)\n+- Better handle invalid import name.\n+ (`#441 <https://github.com/vivisect/vivisect/pull/441>`_)\n+\nv1.0.3 - 2021-05-02\n===================\n"
},
{
"change_type": "MODIFY",
"old_path": "setup.py",
"new_path": "setup.py",
"diff": "from setuptools import find_packages, setup\nfrom os import path\n-VERSION = '1.0.3'\n+VERSION = '1.0.4'\ndirn = path.abspath(path.dirname(__file__))\nwith open(path.join(dirn, 'README.md'), 'r') as fd:\n"
},
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -2293,6 +2293,6 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these yourself\n-version = (1, 0, 3)\n+version = (1, 0, 4)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -3135,6 +3135,6 @@ def getVivPath(*pathents):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these directly\n-version = (1, 0, 3)\n+version = (1, 0, 4)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
V104 release prep (#442)
Prep for releasing v1.0.4
|
718,768 |
25.08.2021 10:17:50
| 21,600 |
fea0883f8368c339eec1050aaa8e7d2108164b98
|
envi: memory: return None on truncated ptr read
closes
|
[
{
"change_type": "MODIFY",
"old_path": "envi/memory.py",
"new_path": "envi/memory.py",
"diff": "import re\nimport struct\n+import logging\nimport envi\nimport envi.exc as e_exc\n@@ -11,6 +12,10 @@ A module containing memory utilities and the definition of the\nmemory access API used by all vtoys trace/emulators/workspaces.\n\"\"\"\n+\n+logger = logging.getLogger(__name__)\n+\n+\ndef getPermName(perm):\n'''\nReturn the human readable name for a *single* memory\n@@ -167,7 +172,8 @@ class IMemory:\n# FIXME change this (and all uses of it) to passing in format...\nif len(bytes) != size:\n- raise Exception(\"Read gave wrong length at va: 0x%.8x (wanted %d got %d)\" % (addr, size, len(bytes)))\n+ logger.warning(\"Read gave wrong length at va: 0x%.8x (wanted %d got %d)\", addr, size, len(bytes))\n+ return None\nreturn e_bits.parsebytes(bytes, 0, size, False, self.getEndian())\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
envi: memory: return None on truncated ptr read (#444)
closes #443
|
718,776 |
01.09.2021 05:59:53
| -7,200 |
27ea43a1ea34dfeb866d755608a2e4f1e876bbba
|
only do rep on string instructions
* only do rep on string instructions
closes
* use REP opcodes tuple
|
[
{
"change_type": "MODIFY",
"old_path": "envi/archs/i386/emu.py",
"new_path": "envi/archs/i386/emu.py",
"diff": "@@ -9,7 +9,7 @@ from envi.const import *\nimport envi.exc as e_exc\nimport envi.bits as e_bits\n-from envi.archs.i386.opconst import PREFIX_REX_W\n+from envi.archs.i386.opconst import PREFIX_REX_W, REP_OPCODES\nfrom envi.archs.i386.regs import *\nfrom envi.archs.i386.disasm import *\nfrom envi.archs.i386 import i386Module\n@@ -245,8 +245,9 @@ class IntelEmulator(i386RegisterContext, envi.Emulator):\nif meth is None:\nraise e_exc.UnsupportedInstruction(self, op)\n+ # The behavior of the REP prefix is undefined when used with non-string instructions.\nrep_prefix = op.prefixes & PREFIX_REP_MASK\n- if rep_prefix and not self.getEmuOpt('i386:reponce'):\n+ if rep_prefix and op.opcode in REP_OPCODES and not self.getEmuOpt('i386:reponce'):\n# REP instructions (REP/REPNZ/REPZ/REPSIMD) get their own handlers\nhandler = self.__rep_prefix_handlers__.get(rep_prefix)\nnewpc = handler(meth, op)\n@@ -273,9 +274,6 @@ class IntelEmulator(i386RegisterContext, envi.Emulator):\nThen the instruction is repeated and ECX decremented until either\nECX reaches 0 or the ZF is cleared.\n'''\n- if op.mnem.startswith('nop'):\n- return\n-\necx = emu.getRegister(REG_ECX)\nemu.setFlag(EFLAGS_ZF, 1)\n"
},
{
"change_type": "MODIFY",
"old_path": "envi/archs/i386/opconst.py",
"new_path": "envi/archs/i386/opconst.py",
"diff": "@@ -197,15 +197,24 @@ INS_OFLOW = INS_TRAPS | 0x08 # gen overflow trap\n#/* INS_SYSTEM */\nINS_HALT = INS_SYSTEM | 0x01 # halt machine\n-INS_IN = INS_SYSTEM | 0x02 # input form port\n+INS_IN = INS_SYSTEM | 0x02 # input from port\nINS_OUT = INS_SYSTEM | 0x03 # output to port\nINS_CPUID = INS_SYSTEM | 0x04 # iden\nINS_NOP = INS_OTHER | 0x01\nINS_BCDCONV = INS_OTHER | 0x02 # convert to/from BCD\nINS_SZCONV = INS_OTHER | 0x03 # convert size of operand\n-INS_CRYPT = INS_OTHER | 0x4 # AES-NI instruction support\n+INS_CRYPT = INS_OTHER | 0x04 # AES-NI instruction support\n+# string instructions that support REP prefix\n+REP_OPCODES = (\n+ INS_IN, # INS\n+ INS_OUT, # OUTS\n+ INS_STRMOV, # MOVS\n+ INS_STRLOAD, # LODS\n+ INS_STRSTOR, # STOS\n+ INS_STRCMP # CMPS, SCAS\n+ )\nOP_R = 0x001\nOP_W = 0x002\n"
},
{
"change_type": "MODIFY",
"old_path": "envi/tests/test_arch_i386_emu.py",
"new_path": "envi/tests/test_arch_i386_emu.py",
"diff": "@@ -108,7 +108,12 @@ i386Tests = [\n# lea ecx,dword [esp + -973695896] (test SIB getOperAddr)\n{'bytes': '8d8c246894f6c5',\n'setup': ({'esp': 0x7fd0}, {}),\n- 'tests': ({'ecx': 0xc5f71438}, {})}\n+ 'tests': ({'ecx': 0xc5f71438}, {})},\n+ # rep ret\n+ # The behavior of the REP prefix is undefined when used with non-string instructions.\n+ {'bytes': 'f3c3',\n+ 'setup': ({'ecx': 0x1}, {'esp': b'\\x00\\x00\\x00\\x60'}),\n+ 'tests': ({'ecx': 0x1, 'eip': 0x60000000}, {})}\n]\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
only do rep on string instructions (#447)
* only do rep on string instructions
closes #446
* use REP opcodes tuple
Co-authored-by: atlas0fd00m <atlas@r4780y.com>
|
718,768 |
14.09.2021 17:46:06
| 21,600 |
bfa65d4be4aac24092434910407ba280cd5e1900
|
elf: use integer division for struct array count
closes
|
[
{
"change_type": "MODIFY",
"old_path": "Elf/__init__.py",
"new_path": "Elf/__init__.py",
"diff": "@@ -422,7 +422,7 @@ class Elf(vs_elf.Elf32, vs_elf.Elf64):\n# only parse the symbols that are not already accounted for.\n# symbols are ordered, so existence of index Y is always the same\nsym = self._cls_symbol(bigend=self.bigend)\n- count = len(symtab) / len(sym)\n+ count = len(symtab) // len(sym)\ndiff = count - len(self.dynamic_symbols)\nif diff == 0:\nreturn\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
elf: use integer division for struct array count (#455)
closes #452
|
718,770 |
14.09.2021 23:08:32
| 14,400 |
0108e9a4331fdf877c60877a7a6c16d37b9e9c40
|
Bugfix: Elf align to page
Bugfix: Align Elf/PE files to page.
|
[
{
"change_type": "MODIFY",
"old_path": "envi/bits.py",
"new_path": "envi/bits.py",
"diff": "@@ -353,3 +353,14 @@ def masktest(s):\ndef domask(testval):\nreturn testval & maskin == matchval\nreturn domask\n+\n+def align(origsize, alignment):\n+ '''\n+ Returns an aligned size based on alignment argument\n+ '''\n+ remainder = origsize % alignment\n+ if remainder == 0:\n+ return origsize\n+ else:\n+ return origsize + (alignment - remainder)\n+\n"
},
{
"change_type": "MODIFY",
"old_path": "envi/memory.py",
"new_path": "envi/memory.py",
"diff": "@@ -137,7 +137,7 @@ class IMemory:\ndef allocateMemory(self, size, perms=MM_RWX, suggestaddr=0):\nraise Exception(\"must implement allocateMemory!\")\n- def addMemoryMap(self, mapva, perms, fname, bytes):\n+ def addMemoryMap(self, mapva, perms, fname, bytes, align=None):\nraise Exception(\"must implement addMemoryMap!\")\ndef getMemoryMaps(self):\n@@ -412,13 +412,13 @@ class MemoryObject(IMemory):\nself._map_defs = []\nself._supervisor = False\n- def allocateMemory(self, size, perms=MM_RWX, suggestaddr=0x1000, name='', fill=b'\\0'):\n+ def allocateMemory(self, size, perms=MM_RWX, suggestaddr=0x1000, name='', fill=b'\\0', align=None):\n'''\nFind a free block of memory (no maps exist) and allocate a new map\nUses findFreeMemoryBlock()\n'''\nbaseva = self.findFreeMemoryBlock(size, suggestaddr)\n- self.addMemoryMap(baseva, perms, name, fill*size)\n+ self.addMemoryMap(baseva, perms, name, fill*size, align)\nreturn baseva\ndef findFreeMemoryBlock(self, size, suggestaddr=0x1000, MIN_MEM_ADDR = 0x1000):\n@@ -468,15 +468,22 @@ class MemoryObject(IMemory):\nreturn baseva\n- def addMemoryMap(self, va, perms, fname, bytez):\n+ def addMemoryMap(self, va, perms, fname, bytez, align=None):\n'''\nAdd a memory map to this object...\n+ Returns the length of the map (since alignment could alter it)\n'''\n+ if align:\n+ curlen = len(bytez)\n+ newlen = e_bits.align(curlen, align)\n+ delta = newlen - curlen\n+ bytez += b'\\x00' * delta\n+\nmsize = len(bytez)\nmmap = (va, msize, perms, fname)\nhlpr = [va, va+msize, mmap, bytez]\nself._map_defs.append(hlpr)\n- return\n+ return msize\ndef delMemoryMap(self, mapva):\n'''\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -1793,12 +1793,16 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nreturn xrefs\nreturn [ xtup for xtup in xrefs if xtup[XR_RTYPE] == rtype ]\n- def addMemoryMap(self, va, perms, fname, bytes):\n+ def addMemoryMap(self, va, perms, fname, bytes, align=None):\n\"\"\"\nAdd a memory map to the workspace. This is the *only* way to\nget memory backings into the workspace.\n\"\"\"\n- self._fireEvent(VWE_ADDMMAP, (va, perms, fname, bytes))\n+ self._fireEvent(VWE_ADDMMAP, (va, perms, fname, bytes, align))\n+\n+ # since we don't return anything from _fireEvent(), pull the new info:\n+ mva, msz, mperm, mbytes = self.getMemoryMap(va)\n+ return msz\ndef delMemoryMap(self, mapva):\n'''\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/base.py",
"new_path": "vivisect/base.py",
"diff": "@@ -383,10 +383,16 @@ class VivWorkspaceCore(viv_impapi.ImportApi):\nself._call_graph.setNodeProp(fnode, 'repr', name)\ndef _handleADDMMAP(self, einfo):\n+ if len(einfo) == 5:\n+ # new \"alignment-friendly\" event\n+ va, perms, fname, mbytes, align = einfo\n+ else:\n+ # DEPRECATED (21-09-13) - old event style, to support older .viv's\nva, perms, fname, mbytes = einfo\n- e_mem.MemoryObject.addMemoryMap(self, va, perms, fname, mbytes)\n+ align = None\n+\n+ blen = e_mem.MemoryObject.addMemoryMap(self, va, perms, fname, mbytes, align)\n- blen = len(mbytes)\nself.locmap.initMapLookup(va, blen)\nself.blockmap.initMapLookup(va, blen)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -6,6 +6,7 @@ import collections\nimport Elf\nimport envi.bits as e_bits\n+import envi.const as e_const\nimport vivisect\nimport vivisect.parsers as v_parsers\n@@ -214,11 +215,11 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\ncontinue\nlogger.info('Loading: %s', pgm)\nbytez = elf.readAtOffset(pgm.p_offset, pgm.p_filesz)\n- bytez += b\"\\x00\" * (pgm.p_memsz - pgm.p_filesz)\n+ bytez += b'\\x00' * (pgm.p_memsz - pgm.p_filesz)\npva = pgm.p_vaddr\nif addbase:\npva += baseaddr\n- vw.addMemoryMap(pva, pgm.p_flags & 0x7, fname, bytez)\n+ vw.addMemoryMap(pva, pgm.p_flags & 0x7, fname, bytez, align=e_const.PAGE_SIZE)\nelse:\nlogger.info('Skipping: %s', pgm)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/pe.py",
"new_path": "vivisect/parsers/pe.py",
"diff": "@@ -32,14 +32,6 @@ for mod in (PE, vtrace):\n# 0x183 DEC Alpha AXP\n-def align(v, alignment):\n- remainder = v % alignment\n- if remainder == 0:\n- return v\n- else:\n- return v + (alignment - remainder)\n-\n-\ndef parseFile(vw, filename, baseaddr=None):\npe = PE.PE(open(filename, \"rb\"))\nreturn loadPeIntoWorkspace(vw, pe, filename=filename, baseaddr=baseaddr)\n@@ -327,21 +319,12 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\n# According to http://code.google.com/p/corkami/wiki/PE#section_table if SizeOfRawData is larger than VirtualSize, VS is used..\nreadsize = sec.SizeOfRawData if sec.SizeOfRawData < sec.VirtualSize else sec.VirtualSize\n- # align the read to the FileAlignment\n- readsize = align(readsize, filealign)\n-\nsecoff = pe.rvaToOffset(secrva)\nsecbytes = pe.readAtOffset(secoff, readsize, shortok=True)\nslen = len(secbytes)\n- if slen != readsize:\n- logger.warning(\"Section at offset 0x%x should have 0x%x bytes, but we only got 0x%x bytes\", secoff, readsize, slen)\n+ secbytes += b'\\x00' * (sec.VirtualSize - slen)\n- if slen != align(sec.VirtualSize, secalign):\n- # pad the section up to next the SectionAlignment\n- secbytes += b'\\x00' * (align(sec.VirtualSize, secalign) - slen)\n-\n- slen = len(secbytes)\n- vw.addMemoryMap(secbase, mapflags, fname, secbytes)\n+ slen = vw.addMemoryMap(secbase, mapflags, fname, secbytes, align=filealign)\nvw.addSegment(secbase, slen, secname, fname)\n# Mark dead data on resource and import data directories\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/storage/mpfile.py",
"new_path": "vivisect/storage/mpfile.py",
"diff": "@@ -34,7 +34,7 @@ def vivEventsToFile(filename, events):\nfor event in events:\nif event[0] == 20:\nmape = base64.b64encode(event[1][3])\n- event = (event[0], (event[1][0], event[1][1], event[1][2], mape))\n+ event = (event[0], (event[1][0], event[1][1], event[1][2], mape, event[1][4]))\nmsgpack.pack(event, f, use_bin_type=False)\n@@ -54,6 +54,10 @@ def vivEventsFromFile(filename):\nfor event in unpacker:\nif event[0] == 20:\nmape = base64.b64decode(event[1][3])\n+ if len(event[1]) == 5:\n+ event = (event[0], (event[1][0], event[1][1], event[1][2], mape, event[1][4]))\n+ else:\n+ # DEPRECATED: for loading older MPFILEs\nevent = (event[0], (event[1][0], event[1][1], event[1][2], mape))\nevents.append(event)\nreturn events\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/stabilitydata/redxor.stability.json",
"new_path": "vivisect/tests/stabilitydata/redxor.stability.json",
"diff": "4,\nnull\n],\n+ [\n+ 4234248,\n+ 8,\n+ 4,\n+ null\n+ ],\n[\n4229536,\n8,\n[]\n],\n[\n- 4231367,\n- 19,\n+ 4231882,\n+ 5,\n2,\n[]\n],\n[\n- 4231882,\n- 5,\n+ 4231367,\n+ 19,\n2,\n[]\n],\n4228375,\n\"redxor.CheckLKM\"\n],\n+ [\n+ 6336456,\n+ \"_end\"\n+ ],\n[\n6334184,\n\"redxor.coOutputWPos\"\n4230600,\n\"str_HTTP/1.0\\rUser-Ag_00408dc8\"\n],\n- [\n- 4231367,\n- \"str_out of pty devic_004090c7\"\n- ],\n[\n4231882,\n\"str_%s_004092ca\"\n],\n+ [\n+ 4231367,\n+ \"str_out of pty devic_004090c7\"\n+ ],\n[\n4231887,\n\"str_%s/.bash_profile_004092cf\"\n1,\n1\n],\n+ [\n+ 4234248,\n+ 0,\n+ 3,\n+ 0\n+ ],\n[\n4229536,\n0,\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testmalware.py",
"new_path": "vivisect/tests/testmalware.py",
"diff": "@@ -28,10 +28,10 @@ class MalwareTests(unittest.TestCase):\nvw = self.mmap_vw\nmaps = set([\n(268435456, 0x1000, 4, '0e25aa791c9119108af073bc9e9d0fa2'),\n- (268439552, 0x67000, 5, '0e25aa791c9119108af073bc9e9d0fa2'),\n- (268861440, 0xe000, 4, '0e25aa791c9119108af073bc9e9d0fa2'),\n- (268918784, 0x14000, 6, '0e25aa791c9119108af073bc9e9d0fa2'),\n- (269004800, 0x5000, 4, '0e25aa791c9119108af073bc9e9d0fa2'),\n+ (268439552, 0x66c00, 5, '0e25aa791c9119108af073bc9e9d0fa2'),\n+ (268861440, 0xdc00, 4, '0e25aa791c9119108af073bc9e9d0fa2'),\n+ (268918784, 0x13600, 6, '0e25aa791c9119108af073bc9e9d0fa2'),\n+ (269004800, 0x4600, 4, '0e25aa791c9119108af073bc9e9d0fa2'),\n])\nself.assertEqual(maps, set(vw.getMemoryMaps()))\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testpe.py",
"new_path": "vivisect/tests/testpe.py",
"diff": "@@ -535,10 +535,10 @@ class PETests(unittest.TestCase):\nans = {\n# name -> (Base, Size, Flags)\n'PE_Header': (0x400000, 0x1000, e_memory.MM_READ),\n- '.text': (0x401000, 0x72000, e_memory.MM_READ | e_memory.MM_EXEC),\n- '.data': (0x4b6000, 0x5000, e_memory.MM_READ | e_memory.MM_WRITE),\n- '.rdata': (0x473000, 0x43000, e_memory.MM_READ),\n- '.reloc': (0x4bf000, 0x7000, e_memory.MM_READ),\n+ '.text': (0x401000, 0x71800, e_memory.MM_READ | e_memory.MM_EXEC),\n+ '.data': (0x4b6000, 0x4200, e_memory.MM_READ | e_memory.MM_WRITE),\n+ '.rdata': (0x473000, 0x42e00, e_memory.MM_READ),\n+ '.reloc': (0x4bf000, 0x6800, e_memory.MM_READ),\n}\nfor sva, ssize, sname, sfname in vw.getSegments():\nself.assertEqual(ans[sname][0], sva)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -463,17 +463,17 @@ class VivisectTest(unittest.TestCase):\nself.assertTrue(len(vw.getLocations()) > 76000)\n# tuples are Name, Number of Locations, Size in bytes, Percentage of space\n- ans = {0: ('Undefined', 0, 70828, 18),\n+ ans = {0: ('Undefined', 0, 55596, 14),\n1: ('Num/Int', 715, 3703, 0),\n2: ('String', 265, 6485, 1),\n3: ('Unicode', 174, 5596, 1),\n4: ('Pointer', 361, 2888, 0),\n- 5: ('Opcode', 72507, 279377, 71),\n+ 5: ('Opcode', 72507, 279377, 73),\n6: ('Structure', 1018, 12740, 3),\n7: ('Clsid', 0, 0, 0),\n8: ('VFTable', 0, 0, 0),\n9: ('Import Entry', 370, 2960, 0),\n- 10: ('Pad', 865, 8639, 2)}\n+ 10: ('Pad', 864, 8511, 2)}\ndist = vw.getLocationDistribution()\nfor loctype, locdist in dist.items():\nself.assertEqual(locdist, ans[loctype])\n@@ -1143,12 +1143,12 @@ class VivisectTest(unittest.TestCase):\n'PE_Header': (0x140000000, 0x1000, e_memory.MM_READ),\n'.text': (0x140001000, 0x49000, e_memory.MM_READ | e_memory.MM_EXEC),\n'.rdata': (0x14004a000, 0xc000, e_memory.MM_READ),\n- '.data': (0x140056000, 0x3000, e_memory.MM_READ | e_memory.MM_WRITE),\n+ '.data': (0x140056000, 0x2a00, e_memory.MM_READ | e_memory.MM_WRITE),\n'.pdata': (0x140059000, 0x3000, e_memory.MM_READ),\n- '.00cfg': (0x14005c000, 0x1000, e_memory.MM_READ),\n- '.freestd': (0x14005d000, 0x1000, e_memory.MM_READ),\n- '.tls': (0x14005e000, 0x1000, e_memory.MM_READ | e_memory.MM_WRITE),\n- '.reloc': (0x140092000, 0x1000, e_memory.MM_READ),\n+ '.00cfg': (0x14005c000, 0x200, e_memory.MM_READ),\n+ '.freestd': (0x14005d000, 0x200, e_memory.MM_READ),\n+ '.tls': (0x14005e000, 0x200, e_memory.MM_READ | e_memory.MM_WRITE),\n+ '.reloc': (0x140092000, 0x400, e_memory.MM_READ),\n}\nfor sva, ssize, sname, sfname in vw.getSegments():\nself.assertEqual(ans[sname][0], sva)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Bugfix: Elf align to page (#451)
Bugfix: Align Elf/PE files to page.
|
718,768 |
15.09.2021 10:32:41
| 21,600 |
08e7a937d2729fb433286e97dd7dbc778cbf1f4e
|
elfplt_late: log to appropriate logger, not root
closes
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/elfplt_late.py",
"new_path": "vivisect/analysis/elf/elfplt_late.py",
"diff": "@@ -106,14 +106,14 @@ def analyzePLT(vw, pltva, pltsz):\nfillPLTviaGOTXrefs(vw, jmpheur, pltva, pltsz)\nelse:\n- logging.info(\"analyzePLT(0x%x, 0x%x) skipping GOT-XREF-Offset method: no existing functions found\", pltva, pltsz)\n+ logger.info(\"analyzePLT(0x%x, 0x%x) skipping GOT-XREF-Offset method: no existing functions found\", pltva, pltsz)\n# PLT-Func-Distance method\nif len(distanceheur):\nfillPLTGaps(vw, curplts, distanceheur, pltva, pltsz)\nelse:\n- logging.info(\"skipping analyzePLT(0x%x, 0x%x) (PLT-Func-Distance method): no existing functions found\", pltva, pltsz)\n+ logger.info(\"skipping analyzePLT(0x%x, 0x%x) (PLT-Func-Distance method): no existing functions found\", pltva, pltsz)\nlogger.info(\"elfplt_late (done): pltva: 0x%x, %d\", pltva, pltsz)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
elfplt_late: log to appropriate logger, not root (#458)
closes #457
|
718,765 |
17.09.2021 11:51:31
| 14,400 |
d5b895eeeb19a0304d965cbf4fc86a5f3a5183c5
|
AddRelocation Robustness
Add some safety to relocs in elf parsing.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -389,11 +389,16 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nExpects data to have whatever is necessary for the reloc type. eg. addend\n\"\"\"\n# split \"current\" va into fname and offset. future relocations will want to base all va's from an image base\n- mmva, mmsz, mmperm, fname = self.getMemoryMap(va) # FIXME: getFileByVa does not obey file defs\n+ mmap = self.getMemoryMap(va)\n+ if not mmap:\n+ logger.warning('addRelocation: No matching map found for %s', va)\n+ return None\n+ mmva, mmsz, mmperm, fname = mmap # FIXME: getFileByVa does not obey file defs\nimgbase = self.getFileMeta(fname, 'imagebase')\noffset = va - imgbase\nself._fireEvent(VWE_ADDRELOC, (fname, offset, rtype, data))\n+ return self.getRelocation(va)\ndef getRelocations(self):\n\"\"\"\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -639,7 +639,8 @@ def applyRelocs(elf, vw, addbase=False, baseaddr=0):\nif rtype == Elf.R_X86_64_IRELATIVE:\n# before making import, let's fix up the pointer as a BASEPTR Relocation\nptr = r.r_addend\n- vw.addRelocation(rlva, RTYPE_BASEPTR, ptr)\n+ rloc = vw.addRelocation(rlva, RTYPE_BASEPTR, ptr)\n+ if rloc:\nlogger.info('Reloc: R_X86_64_IRELATIVE 0x%x', rlva)\nif rtype in (Elf.R_386_JMP_SLOT, Elf.R_X86_64_GLOB_DAT, Elf.R_X86_64_IRELATIVE):\n@@ -671,7 +672,9 @@ def applyRelocs(elf, vw, addbase=False, baseaddr=0):\n# first make it a relocation that is based on the imagebase\nptr = r.r_addend\nlogger.info('R_X86_64_IRELATIVE: adding Relocation 0x%x -> 0x%x (name: %r %r) ', rlva, ptr, name, dmglname)\n- vw.addRelocation(rlva, RTYPE_BASEPTR, ptr)\n+ rloc = vw.addRelocation(rlva, RTYPE_BASEPTR, ptr)\n+ if rloc is not None:\n+ continue\n# next get the target and find a name, since the reloc itself doesn't have one\ntgt = vw.readMemoryPtr(rlva)\n@@ -741,6 +744,7 @@ def applyRelocs(elf, vw, addbase=False, baseaddr=0):\n# that does *not* mean it's not an IMPORT\nif ptr and not isPLT(vw, ptr):\nlogger.info('R_ARM_JUMP_SLOT: adding Relocation 0x%x -> 0x%x (%s) ', rlva, ptr, dmglname)\n+ # even if addRelocation fails, still make the name, same thing down in GLOB_DAT\nif addbase:\nvw.addRelocation(rlva, vivisect.RTYPE_BASEPTR, ptr)\nelse:\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -584,6 +584,14 @@ class VivisectTest(unittest.TestCase):\nfor va, disp in ans:\nself.assertEqual(disp, vw.reprVa(va))\n+ def test_basic_reloc(self):\n+ vw = self.chgrp_vw\n+ rloc = vw.addRelocation(0xdeadbeef, v_const.RTYPE_BASEOFF)\n+ self.assertIsNone(rloc)\n+\n+ rloc = vw.addRelocation(0x08054184, v_const.RTYPE_BASEPTR)\n+ self.assertIsNotNone(rloc)\n+\ndef test_naughty(self):\n'''\nTest us some error conditions\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
AddRelocation Robustness (#456)
Add some safety to relocs in elf parsing.
|
718,770 |
17.09.2021 22:45:35
| 14,400 |
5c3f69977f433c5613ceaec57763644dd919ac27
|
allow duplicate init_function and fini_function entries that differ between section and dynamics analysis
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -352,15 +352,19 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nvw.makeString(sva)\nelif sname == \".init\":\n- vw.makeName(sva, \"init_function\", filelocal=True)\n- new_functions.append((\"init_function\", sva))\n+ init_tup = (\"init_function\", sva)\n+ if init_tup not in new_functions:\n+ vw.makeName(sva, \"init_function\", filelocal=True, makeuniq=True)\n+ new_functions.append(init_tup)\nelif sname == \".init_array\":\nmakeFunctionTable(elf, vw, sec.sh_addr, size, 'init_function', new_functions, new_pointers, baseaddr, addbase)\nelif sname == \".fini\":\n- vw.makeName(sva, \"fini_function\", filelocal=True)\n- new_functions.append((\"fini_function\", sva))\n+ fini_tup = (\"fini_function\", sva)\n+ if fini_tup not in new_functions:\n+ vw.makeName(sva, \"fini_function\", filelocal=True, makeuniq=True)\n+ new_functions.append(fini_tup)\nelif sname == \".fini_array\":\nmakeFunctionTable(elf, vw, sec.sh_addr, size, 'fini_function', new_functions, new_pointers, baseaddr, addbase)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
allow duplicate init_function and fini_function entries that differ between section and dynamics analysis (#459)
|
718,770 |
20.09.2021 22:02:06
| 14,400 |
93a0644e1dd9c5167547b7370fe2f68cec08bab2
|
Vtrace Symbol Test
Add Vtrace Symbol Test
|
[
{
"change_type": "ADD",
"old_path": null,
"new_path": "vtrace/tests/test_symcache.py",
"diff": "+import os\n+import unittest\n+import vtrace.tests as vt_tests\n+\n+\n+class VtraceSymbolTest(vt_tests.VtraceProcessTest):\n+\n+ def test_symbol(self):\n+ symlist = self.trace.getSymList()\n+ # currently there are 6355, but we don't care about exactness, just that\n+ # it's a largish number\n+ self.assertGreater(len(symlist), 1000)\n+ sym0 = symlist[0]\n+\n+ symPyRevType = self.trace.getSymByName('PyReversed_Type')\n+ self.assertIsNotNone(symPyRevType)\n+ self.assertIn(symPyRevType.fname, ('python', 'python3', 'python2', 'libpython3'))\n+ self.assertEqual(symPyRevType.name, 'PyReversed_Type')\n+ self.assertAlmostEqual(symPyRevType.size, 0x1a0, delta=20)\n+\n+ symPyRevT2 = self.trace.getSymByAddr(symPyRevType.value)\n+ self.assertEqual(symPyRevType, symPyRevT2)\n+\n+\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Vtrace Symbol Test (#460)
Add Vtrace Symbol Test
|
718,765 |
22.09.2021 12:53:11
| 14,400 |
885076381379ca62a7c4ac6519ddbc79323c661b
|
V1.0.5 Release Prep
v1.0.5 Release Prep
|
[
{
"change_type": "MODIFY",
"old_path": ".bumpversion.cfg",
"new_path": ".bumpversion.cfg",
"diff": "[bumpversion]\n-current_version = 1.0.4\n+current_version = 1.0.5\n[bumpversion:file:setup.py]\nsearch = VERSION = '{current_version}'\n"
},
{
"change_type": "MODIFY",
"old_path": "CHANGELOG.rst",
"new_path": "CHANGELOG.rst",
"diff": "Vivisect Changelog\n******************\n+V1.0.5 - 2021-09-10\n+===================\n+\n+Fixes\n+-----\n+- Fix ascii string size when the string terminates at the end of a memory map.\n+ (`#437 <https://github.com/vivisect/vivisect/pull/437>`_)\n+- Better handle PE delay imports that use VA pointers instead of RVA pointers.\n+ (`#439 <https://github.com/vivisect/vivisect/pull/439>`_)\n+- envi.IMemory.readMemValue: return None on truncated read.\n+ (`#444 <https://github.com/vivisect/vivisect/pull/444>`_)\n+- Only apply the rep prefix on string instructions in intel emulation.\n+ (`#447 <https://github.com/vivisect/vivisect/pull/447>`_)\n+- Fix a pair of regressions in ELF analysis.\n+ (`#448 <https://github.com/vivisect/vivisect/pull/448>`_)\n+- Align ELF memory maps to page.\n+ (`#451 <https://github.com/vivisect/vivisect/pull/451>`_)\n+- Integer division for struct array count in ELF.\n+ (`#455 <https://github.com/vivisect/vivisect/pull/455>`_)\n+- Safe harness for addRelocation method on the workspace.\n+ (`#456 <https://github.com/vivisect/vivisect/pull/456>`_)\n+- Log to appropriate logger in elfplt late module.\n+ (`#458 <https://github.com/vivisect/vivisect/pull/458>`_)\n+- Allow duplicate init and fini functions in ELF files.\n+ (`#459 <https://github.com/vivisect/vivisect/pull/459>`_)\n+- Add Vtrace Symbol test.\n+ (`#460 <https://github.com/vivisect/vivisect/pull/460>`_)\n+\nv1.0.4 - 2021-08-22\n===================\nFeatures\n--------\n- Add structures to UI and a compressed version of the file to the meta events.\n- (`#396 <https://github.com/vivisect/vivisect/pull/396`_)\n+ (`#396 <https://github.com/vivisect/vivisect/pull/396>`_)\n- Actual documentation!\n(`#400 <https://github.com/vivisect/vivisect/pull/400>`_)\n- Massive ELFPLT overhaul.\n"
},
{
"change_type": "MODIFY",
"old_path": "setup.py",
"new_path": "setup.py",
"diff": "from setuptools import find_packages, setup\nfrom os import path\n-VERSION = '1.0.4'\n+VERSION = '1.0.5'\ndirn = path.abspath(path.dirname(__file__))\nwith open(path.join(dirn, 'README.md'), 'r') as fd:\n"
},
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -2293,6 +2293,6 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these yourself\n-version = (1, 0, 4)\n+version = (1, 0, 5)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -3144,6 +3144,6 @@ def getVivPath(*pathents):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these directly\n-version = (1, 0, 4)\n+version = (1, 0, 5)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
V1.0.5 Release Prep (#449)
v1.0.5 Release Prep
|
718,765 |
23.09.2021 23:56:54
| 14,400 |
91f3952994815e2fa2da04482b70dd882a1eb451
|
add changelog to docs
thanks
|
[
{
"change_type": "MODIFY",
"old_path": "docs/index.rst",
"new_path": "docs/index.rst",
"diff": "@@ -18,6 +18,7 @@ Welcome to vivisect's documentation!\nvivisect/extensions\nvivisect/scripting\nvivisect/remote\n+ vivisect/changelog\nIndices and tables\n"
},
{
"change_type": "ADD",
"old_path": null,
"new_path": "docs/vivisect/changelog.rst",
"diff": "+.. _changelog:\n+\n+.. include:: ../../CHANGELOG.rst\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
add changelog to docs (#462)
thanks @rakuy0
|
718,765 |
24.09.2021 00:27:23
| 14,400 |
0274ff1dee066ccb9c9f5072fb7ac7b98655f854
|
add test for workspace failure
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/teststorage.py",
"new_path": "vivisect/tests/teststorage.py",
"diff": "@@ -114,3 +114,13 @@ class StorageTests(unittest.TestCase):\nbasicfile.close()\nos.unlink(mpfile.name)\nos.unlink(basicfile.name)\n+\n+ def test_bad_event(self):\n+ vw = vivisect.VivWorkspace()\n+ with self.assertLogs() as logcap:\n+ vw.importWorkspace([(VWE_MAX + 1, (0xabad1dea, 4, 3, 'nope')),\n+ (VWE_ADDFILE, ('VivisectFile', 0x1000, '3bfdad02b9a6522c84e356cf8f69135b'))])\n+ files = vw.getFiles()\n+ self.assertIn(\"IndexError: list index out of range\", ''.join(logcap.output))\n+ self.assertEqual(1, len(files))\n+ self.assertEqual('VivisectFile', files[0])\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
add test for workspace failure (#463)
|
718,765 |
19.10.2021 19:33:29
| 14,400 |
a116ffeb695654da5a0949ad6f0c5ba78bea5d18
|
fix some qt issues in vdb
|
[
{
"change_type": "MODIFY",
"old_path": "envi/qt/memorymap.py",
"new_path": "envi/qt/memorymap.py",
"diff": "@@ -10,6 +10,7 @@ import envi.cli as e_cli\nfrom vqt.common import ACT\nimport vqt.tree as vq_tree\n+# TODO: Why is this here and not jut mixed in w/ vtrace since they're the only consumers?\nclass VQMemoryMapView(vq_tree.VQTreeView):\ndef __init__(self, mem, parent=None):\n"
},
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -81,7 +81,7 @@ def setupBreakOnEntry(trace):\notb = vtrace.OneTimeBreak(None, expression=entrySymExpr)\ntrace.addBreakpoint(otb)\n-class VdbTrace:\n+class VdbTrace(object):\n\"\"\"\nUsed to hand thing that need a persistant reference to a trace\nwhen using vdb to manage tracers.\n"
},
{
"change_type": "MODIFY",
"old_path": "vdb/qt/main.py",
"new_path": "vdb/qt/main.py",
"diff": "@@ -220,6 +220,9 @@ class VdbWindow(vq_app.VQMainCmdWindow):\nself.addHotKey('ctrl+b','debug:break')\nself.addHotKey('ctrl+p','vdb:view:python')\n+ self.addHotKey('ctrl+w', 'file:quit')\n+ self.addHotKeyTarget('file:quit', self.close)\n+\n# Get hotkey overrides\nself.loadHotKeys(self._vq_settings)\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/qt.py",
"new_path": "vtrace/qt.py",
"diff": "@@ -71,7 +71,7 @@ class RegisterListModel(envi_qt_memory.EnviNavModel):\nreturn True\n-class RegistersListView(vq_tree.VQTreeView, VQTraceNotifier):\n+class RegistersListView(VQTraceNotifier, vq_tree.VQTreeView):\n'''\nA pure \"list view\" object for registers\n'''\n@@ -199,12 +199,12 @@ class RegistersView(QWidget):\n# show general in dropdown by default if exists, otherwise all\n# (preferences will re-set)\nif 'general' in self.regviews:\n- self.regViewNameSelected('general')\nidx = self.viewnames.findText('general')\n+ self.regViewNameSelected('general')\nself.viewnames.setCurrentIndex(idx)\nelse:\n- self.regViewNameSelected('all')\nidx = self.viewnames.findText('all')\n+ self.regViewNameSelected('all')\nself.viewnames.setCurrentIndex(idx)\nstatusreg_widget = VQFlagsGridView(trace=trace, parent=self)\n@@ -216,8 +216,7 @@ class RegistersView(QWidget):\nsplitview.addWidget(statusreg_widget)\nvbox.addWidget(splitview)\n- sig = QtCore.SIGNAL('currentIndexChanged(str)')\n- self.viewnames.connect(self.viewnames, sig, self.regViewNameSelected)\n+ self.viewnames.currentTextChanged.connect(self.regViewNameSelected)\nself.setLayout(vbox)\n@@ -225,7 +224,7 @@ class RegistersView(QWidget):\nself.reglist.regnames = self.regviews.get(str(name), None)\nself.reglist.vqLoad()\n-class VQFlagsGridView(QWidget, VQTraceNotifier):\n+class VQFlagsGridView(VQTraceNotifier, QWidget):\n'''\nShow the state of the status register (if available).\n'''\n@@ -366,7 +365,7 @@ def getProcessPid(trace=None, parent=None):\nclass FileDescModel(vq_tree.VQTreeModel):\ncolumns = ('Fd','Type','Name')\n-class VQFileDescView(vq_tree.VQTreeView, VQTraceNotifier):\n+class VQFileDescView(VQTraceNotifier, vq_tree.VQTreeView):\ndef __init__(self, trace, parent=None):\nVQTraceNotifier.__init__(self, trace)\n@@ -490,7 +489,7 @@ class VQTraceToolBar(QToolBar, vtrace.Notifier):\nelse:\nself._updateActions(trace.isAttached(), trace.shouldRunAgain())\n-class VQMemoryMapView(envi_qt_memmap.VQMemoryMapView, VQTraceNotifier):\n+class VQMemoryMapView(VQTraceNotifier, envi_qt_memmap.VQMemoryMapView):\n'''\nA memory map view which is sensitive to the status of a\ntrace object.\n@@ -517,8 +516,8 @@ class VQThreadsView(vq_tree.VQTreeView, VQTraceNotifier):\ndef __init__(self, trace=None, parent=None, selectthread=None):\n# selectthread is an optional endpoint to connect to\n- VQTraceNotifier.__init__(self, trace)\nvq_tree.VQTreeView.__init__(self, parent=parent)\n+ VQTraceNotifier.__init__(self, trace)\nself.setWindowTitle('Threads')\nself.setModel(VQThreadListModel(parent=self))\nself.setAlternatingRowColors(True)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
fix some qt issues in vdb (#466)
|
718,770 |
11.11.2021 17:26:29
| 18,000 |
3dfa3d2ceab45d9e656ee87699b856ca88c62235
|
minor bugfix from Vdb/Viv/Emu bridge pr
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/impemu/platarch/arm.py",
"new_path": "vivisect/impemu/platarch/arm.py",
"diff": "@@ -27,12 +27,7 @@ class ArmWorkspaceEmulator(v_i_emulator.WorkspaceEmulator, e_arm.ArmEmulator):\nva = kwargs.get('va')\nif va:\n- loc = vw.getLocation(va)\n- if loc:\n- lva, lsz, lt, lflags = loc\n- self.setMemArchitecture(lflags)\n- tmode = (lflags & envi.ARCH_MASK == envi.ARCH_THUMB)\n- self.setThumbMode(tmode)\n+ self._prep(va)\ndef setThumbMode(self, thumb=1):\ne_arm.ArmEmulator.setThumbMode(self, thumb)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
minor bugfix from Vdb/Viv/Emu bridge pr
|
718,770 |
11.11.2021 17:26:50
| 18,000 |
06a4484c4f5379c307b930097308dfa69db35c48
|
adding a new posix noret
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -166,6 +166,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nvw.addNoReturnApi(\"*.__assert_fail\")\nvw.addNoReturnApi(\"*.__stack_chk_fail\")\nvw.addNoReturnApi(\"*.pthread_exit\")\n+ vw.addNoReturnApi(\"*.longjmp\")\n# for VivWorkspace, MSB==1, LSB==0... which is the same as True/False\nvw.setEndian(elf.getEndian())\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
adding a new posix noret
|
718,770 |
11.11.2021 17:27:31
| 18,000 |
10fbbd62ea3c4c62d078df641d2cc014702524e3
|
minor exception handling
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/elfplt.py",
"new_path": "vivisect/analysis/elf/elfplt.py",
"diff": "@@ -299,7 +299,6 @@ def analyzePLT(vw, ssva, ssize):\nlogger.error('analyzePLT(0x%x, %r): %s', ssva, ssize, str(e))\n-\nMAX_OPS = 10\ndef analyzeFunction(vw, funcva):\n@@ -386,7 +385,10 @@ def analyzeFunction(vw, funcva):\nelif ltype == vivisect.LOC_POINTER:\n# we have a deref to a pointer.\nfuncname = vw.getName(ltinfo)\n+ if ltinfo:\nlogger.debug(\"0x%x: (0x%x->0x%x) LOC_POINTER by BR_DEREF %r\", funcva, opval, ltinfo, funcname)\n+ else:\n+ logger.debug(\"0x%x: (0x%x->%r) LOC_POINTER by BR_DEREF %r\", funcva, opval, ltinfo, funcname)\nelse:\nlogger.warning(\"0x%x: (0x%x) not LOC_IMPORT or LOC_POINTER?? by BR_DEREF %r\", funcva, opval, loctup)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
minor exception handling
|
718,770 |
11.11.2021 18:14:23
| 18,000 |
104e8c5c0c9b5daacb6bb09b4eb43746aba2bd78
|
bugfix: make elfplt analysis use makeFunctionThunk() in a way that actually catches norets using the fmcb_Thunk() analysis hook.
bugfix: redo how pointers are incremented
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -944,7 +944,8 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nloctup = self.getLocation(va)\nif loctup is not None:\n- offset += loctup[L_SIZE]\n+ nextva = loctup[L_VA] + loctup[L_SIZE]\n+ offset = nextva - mva\nif offset % align:\noffset += align\noffset &= -align\n@@ -1672,7 +1673,7 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nret = []\nreturn ret\n- def makeFunctionThunk(self, fva, thname, addVa=True, filelocal=False):\n+ def makeFunctionThunk(self, fva, thname, addVa=True, filelocal=False, basename=None):\n\"\"\"\nInform the workspace that a given function is considered a \"thunk\" to another.\nThis allows the workspace to process argument inheritance and several other things.\n@@ -1681,13 +1682,14 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\n\"\"\"\nself.checkNoRetApi(thname, fva)\nself.setFunctionMeta(fva, \"Thunk\", thname)\n- n = self.getName(fva)\n- base = thname.split(\".\")[-1]\n+ if basename is None:\n+ basename = thname.split(\".\")[-1]\n+\nif addVa:\n- name = \"%s_%.8x\" % (base,fva)\n+ name = \"%s_%.8x\" % (basename, fva)\nelse:\n- name = base\n+ name = basename\nnewname = self.makeName(fva, name, filelocal=filelocal, makeuniq=True)\napi = self.getImpApi(thname)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/elfplt.py",
"new_path": "vivisect/analysis/elf/elfplt.py",
"diff": "@@ -477,7 +477,8 @@ def analyzeFunction(vw, funcva):\nfuncname = funcname[:-9]\nlogger.info('makeFunctionThunk(0x%x, \"plt_%s\")', funcva, funcname)\n- vw.makeFunctionThunk(funcva, \"plt_\" + funcname, addVa=False, filelocal=True)\n+ vw.makeFunctionThunk(funcva, \"*.\" + funcname, addVa=False, filelocal=True,\n+ basename=\"plt_\" + funcname)\n'''\ndef printPLTs(vw):\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
bugfix: make elfplt analysis use makeFunctionThunk() in a way that actually catches norets using the fmcb_Thunk() analysis hook.
bugfix: redo how pointers are incremented
|
718,770 |
11.11.2021 18:15:12
| 18,000 |
4aef879ff89d7ae1d1131aff053a2b8bf02db0a6
|
cleanup: remove "thread" from the logging format. that should save a few kb each run :)
|
[
{
"change_type": "MODIFY",
"old_path": "envi/common.py",
"new_path": "envi/common.py",
"diff": "@@ -2,7 +2,7 @@ import logging\nimport binascii\nLOG_FORMAT = '%(asctime)s:%(levelname)s:%(name)s: %(message)s' \\\n- '[%(filename)s:%(funcName)s:%(lineno)s:%(threadName)s]'\n+ '[%(filename)s:%(funcName)s:%(lineno)s]'\nEMULOG = 9\nSHITE = 8\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
cleanup: remove "thread" from the logging format. that should save a few kb each run :)
|
718,770 |
11.11.2021 18:16:43
| 18,000 |
abf7082dcf24aa568bb712bc115f0751fc55d2c9
|
make function headers display "NORET" for functions that have been categorized as such.
|
[
{
"change_type": "MODIFY",
"old_path": "envi/qt/html.py",
"new_path": "envi/qt/html.py",
"diff": "@@ -75,6 +75,16 @@ div.codeblock:hover {\nbackground-color: #ffff00;\n}\n+.envi-funcflags {\n+ color: #ff00dd;\n+ background-color: #000000;\n+}\n+\n+.envi-funcflags-selected {\n+ color: #000000;\n+ background-color: #ff00dd;\n+}\n+\n</style>\n<style type=\"text/css\" id=\"cmapstyle\">\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/renderers/__init__.py",
"new_path": "vivisect/renderers/__init__.py",
"diff": "@@ -101,7 +101,12 @@ class WorkspaceRenderer(e_canvas.MemoryRenderer):\n# FIXME color code and get args parsing goin on\nmcanv.addText(\" \")\nxrtag = mcanv.getTag(\"xrefs\")\n- mcanv.addText(\"[%d XREFS]\\n\" % xrcount, tag=xrtag)\n+ mcanv.addText(\"[%d XREFS] \" % xrcount, tag=xrtag)\n+\n+ if self.vw.getMeta('NoReturnApisVa', {}).get(lva):\n+ mcanv.addNameText('NORET', typename='funcflags')\n+\n+ mcanv.addText('\\n')\nmcanv.addText(linepre, tag=vatag)\nmcanv.addText('\\n')\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
make function headers display "NORET" for functions that have been categorized as such.
|
718,770 |
12.11.2021 18:05:50
| 18,000 |
6644f2c255f40de198cc9b4166ade21282dfa396
|
tweak testelf (yay, no longer cares about "unique" names like foo_0 and foo_1)
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testelf.py",
"new_path": "vivisect/tests/testelf.py",
"diff": "@@ -37,12 +37,12 @@ def cmpnames(x, y):\nnmset = [x, y]\nfor nmidx in range(2):\n- name = nmset[nmidx]\n+ va, name = nmset[nmidx]\nif len(name) <= 2:\ncontinue\n- if name[-2:] in (b'_0', b'_1', b'_2', b'_3', b'_4', b'_5', b'_6', b'_7', b'_8', b'_9'):\n- nmset[nmidx] = name[:-2]\n+ if name[-2:] in ('_0', '_1', '_2', '_3', '_4', '_5', '_6', '_7', '_8', '_9'):\n+ nmset[nmidx] = va, name[:-2]\nreturn nmset[0] == nmset[1]\n@@ -156,7 +156,7 @@ class ELFTests(unittest.TestCase):\n# So this portion is because on windows, there's no good python equivalent for cxxfilt that I\n# can find. So we have to skip the portions of the tests that rely on decoding the names\ncmpr = lambda x, y: x == y\n- if platform.system().lower() == 'windows' and testname in comparators:\n+ if testname in comparators:\ncmpr = comparators[testname]\nfor base in baseline:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
tweak testelf (yay, no longer cares about "unique" names like foo_0 and foo_1)
|
718,770 |
12.11.2021 18:08:20
| 18,000 |
b7423c2056d08c7d4c12208c78ff52cc8d0a42c3
|
better context for pointer log messages
more testelf
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/generic/pointers.py",
"new_path": "vivisect/analysis/generic/pointers.py",
"diff": "@@ -24,7 +24,7 @@ def analyze(vw):\ncontinue\nfor xfr, xto, xtype, xinfo in vw.getXrefsFrom(rva):\n- logger.debug('pointer(1): 0x%x -> 0x%x', xfr, xto)\n+ logger.debug('pointer(xref): 0x%x -> 0x%x', xfr, xto)\nvw.analyzePointer(xto)\ndone[xfr] = xto\n@@ -32,9 +32,9 @@ def analyze(vw):\nfor pva, tva, fname, pname in vw.getVaSetRows('PointersFromFile'):\nif vw.getLocation(pva) is None:\nif tva is None:\n- logger.debug('making pointer(2) 0x%x (%r)', pva, pname)\n+ logger.debug('making pointer(fromFile) 0x%x (%r)', pva, pname)\nelse:\n- logger.debug('making pointer(2) 0x%x -> 0x%x (%r)', pva, tva, pname)\n+ logger.debug('making pointer(fromFile) 0x%x -> 0x%x (%r)', pva, tva, pname)\nvw.makePointer(pva, tva, follow=True)\ndone[pva] = tva\n@@ -51,7 +51,7 @@ def analyze(vw):\nlogger.debug('following previously discovered pointer 0x%x -> 0x%x', lva, tva)\ntry:\n- logger.debug('pointer(3): 0x%x -> 0x%x', lva, tva)\n+ logger.debug('pointer: 0x%x -> 0x%x', lva, tva)\nvw.followPointer(tva)\ndone[lva] = tva\nexcept Exception as e:\n@@ -62,7 +62,7 @@ def analyze(vw):\nif vw.isDeadData(pval):\ncontinue\ntry:\n- logger.debug('pointer(4): 0x%x -> 0x%x', addr, pval)\n+ logger.debug('make pointer(found): 0x%x -> 0x%x', addr, pval)\nvw.makePointer(addr, follow=True)\ndone[addr] = pval\nexcept Exception as e:\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testelf.py",
"new_path": "vivisect/tests/testelf.py",
"diff": "@@ -34,6 +34,10 @@ def do_analyze(vw):\nlogging.warning(\"ERROR in analysis module: (%r): %r\", mod, e, exc_info=1)\ndef cmpnames(x, y):\n+ '''\n+ Names comparator. Skips \"_#\" duplicate name parts so that processing order\n+ doesn't matter.\n+ '''\nnmset = [x, y]\nfor nmidx in range(2):\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
better context for pointer log messages
more testelf
|
718,770 |
15.11.2021 00:06:23
| 18,000 |
698ef925da7c805fe440decb14c7d8778334375e
|
tweaks per
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -1678,6 +1678,12 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nInform the workspace that a given function is considered a \"thunk\" to another.\nThis allows the workspace to process argument inheritance and several other things.\n+ If basename is provided, that name is used to create the Vivisect name for the thunk.\n+ If basename is not provided, thname is chopped and used for the Vivisect name.\n+ This difference allows, for example, the Elf loader to make PLT functions named \"plt_<foo>\"\n+ but still use the official thunk name \"*.<foo>\". This thunk name is used to look up\n+ the import api. These \"*.<foo>\" thunk names are also used in the addNoReturnApi().\n+\nUsage: vw.makeFunctionThunk(0xvavavava, \"kernel32.CreateProcessA\")\n\"\"\"\nself.checkNoRetApi(thname, fva)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/impemu/emulator.py",
"new_path": "vivisect/impemu/emulator.py",
"diff": "@@ -496,6 +496,7 @@ class WorkspaceEmulator:\nif self.emustop:\nreturn\n+\niscall = self.checkCall(starteip, endeip, op)\nif self.emustop:\nreturn\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/impemu/platarch/arm.py",
"new_path": "vivisect/impemu/platarch/arm.py",
"diff": "@@ -161,6 +161,7 @@ class ArmWorkspaceEmulator(v_i_emulator.WorkspaceEmulator, e_arm.ArmEmulator):\nvw = self.vw # Save a dereference many many times\nwhile len(todo):\n+\nva, esnap, self.curpath = todo.pop()\nself.setEmuSnap(esnap)\n@@ -193,12 +194,11 @@ class ArmWorkspaceEmulator(v_i_emulator.WorkspaceEmulator, e_arm.ArmEmulator):\nhits[starteip] = h\n# If we ran out of path (branches that went\n- # somewhere that we couldn't follow?\n+ # somewhere that we couldn't follow)?\nif self.curpath is None:\nbreak\ntry:\n-\n# FIXME unify with stepi code...\nop = self.parseOpcode(starteip | tmode)\n@@ -207,8 +207,10 @@ class ArmWorkspaceEmulator(v_i_emulator.WorkspaceEmulator, e_arm.ArmEmulator):\ntry:\nself.emumon.prehook(self, op, starteip)\nexcept v_exc.BadOpBytes as e:\n- logger.debug(repr(e))\n+ logger.debug(str(e))\nbreak\n+ except v_exc.BadOutInstruction:\n+ pass\nexcept Exception as e:\nlogger.log(self._log_level, \"funcva: 0x%x opva: 0x%x: %r (%r) (in emumon prehook: %r)\", funcva, starteip, op, e, self.emumon)\n@@ -224,8 +226,14 @@ class ArmWorkspaceEmulator(v_i_emulator.WorkspaceEmulator, e_arm.ArmEmulator):\nif self.emumon:\ntry:\nself.emumon.posthook(self, op, endeip)\n+ except v_exc.BadOpBytes as e:\n+ logger.debug(str(e))\n+ break\n+ except v_exc.BadOutInstruction:\n+ pass\nexcept Exception as e:\nlogger.log(self._log_level, \"funcva: 0x%x opva: 0x%x: %r (%r) (in emumon posthook: %r)\", funcva, starteip, op, e, self.emumon)\n+\nif self.emustop:\nreturn\n@@ -257,6 +265,15 @@ class ArmWorkspaceEmulator(v_i_emulator.WorkspaceEmulator, e_arm.ArmEmulator):\nvg_path.setNodeProp(self.curpath, 'cleanret', True)\nbreak\n+ # TODO: hook things like error(...) when they have a param that indicates to\n+ # exit. Might be a bit hairy since we'll possibly have to fix up codeblocks\n+ # Make sure we can at least get past the first instruction in certain functions\n+ if self.vw.isNoReturnVa(op.va) and op.va != funcva:\n+ vg_path.setNodeProp(self.curpath, 'cleanret', False)\n+ break\n+\n+ except envi.BadOpcode:\n+ break\nexcept envi.UnsupportedInstruction as e:\nif self.strictops:\nlogger.debug('runFunction breaking after unsupported instruction: 0x%08x %s', e.op.va, e.op.mnem)\n@@ -264,8 +281,10 @@ class ArmWorkspaceEmulator(v_i_emulator.WorkspaceEmulator, e_arm.ArmEmulator):\nelse:\nlogger.debug('runFunction continuing after unsupported instruction: 0x%08x %s', e.op.va, e.op.mnem)\nself.setProgramCounter(e.op.va + e.op.size)\n+ except v_exc.BadOutInstruction:\n+ break\nexcept Exception as e:\n- if self.emumon is not None:\n+ if self.emumon is not None and not isinstance(e, e_exc.BreakpointHit):\nself.emumon.logAnomaly(self, starteip, str(e))\nlogger.debug('runFunction breaking after exception (fva: 0x%x): %s', funcva, e)\nbreak # If we exc during execution, this branch is dead.\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
tweaks per @rakuy0
|
718,770 |
15.11.2021 02:51:23
| 18,000 |
e0c4f65ca4174b63c2ffb7554579d6309132949a
|
lots of updates to support getBaseAddrAndSize() on each file format (except macho, which is a wip)
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2736,6 +2736,11 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nreturn self.normFileName(filename)\nmod = viv_parsers.getParserModule(fmtname)\n+ # get baseaddr and size, then make sure we have a good baseaddr\n+ baseaddr, size = mod.getMemoryBaseAndSize(self, filename=filename, baseaddr=baseaddr)\n+ baseaddr = self.findFreeMemoryBlock(size, baseaddr)\n+ logger.debug(\"loading %r (size: 0x%x) at 0x%x\", filename, size, baseaddr)\n+\nfname = mod.parseFile(self, filename=filename, baseaddr=baseaddr)\nself.initMeta(\"StorageName\", filename+\".viv\")\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/blob.py",
"new_path": "vivisect/parsers/blob.py",
"diff": "@@ -77,3 +77,10 @@ def parseMemory(vw, memobj, baseaddr):\nvw.setFileMeta(fname, 'sha256', v_parsers.sha256Bytes(bytez))\nvw.addMemoryMap(va, perms, fname, bytez)\nvw.setMeta('DefaultCall', archcalls.get(arch,'unknown'))\n+\n+def getBaseAndSize(vw, filename, baseaddr=None):\n+ if baseaddr is None:\n+ baseaddr = vw.config.viv.parsers.blob.baseaddr\n+ size = os.lstat(filename).st_size\n+ return baseaddr, size\n+\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -43,6 +43,87 @@ def parseFd(vw, fd, filename=None, baseaddr=None):\ndef parseMemory(vw, memobj, baseaddr):\nraise Exception('FIXME implement parseMemory for elf!')\n+def getBaseAndSize(vw, filename, baseaddr=None):\n+ '''\n+ Returns the default baseaddr and memory size required to load the file\n+ '''\n+ savebase = baseaddr\n+\n+ fd = open(filename, 'rb')\n+ elf = Elf.Elf(fd)\n+\n+ memmaps = getMemoryMapInfo(elf)\n+ baseaddr = 0xffffffffffffffffffffffff\n+ topmem = 0\n+\n+ for mapva, mperms, mname, mbytes, malign in memmaps:\n+ if mapva < baseaddr:\n+ baseaddr = mapva\n+ endva = mapva + len(mbytes)\n+ if endva > topmem:\n+ topmem = endva\n+\n+ size = topmem - baseaddr\n+ if savebase:\n+ # if we provided a baseaddr, override what the file wants\n+ baseaddr = savebase\n+\n+ return baseaddr, size\n+\n+\n+\n+def getMemoryMapInfo(elf):\n+ '''\n+ Gets the default baseaddr and memory map information\n+ All the information necessary to add memory maps (or get overall size info)\n+ '''\n+ memmaps = []\n+\n+ addbase, baseaddr = getAddBaseAddr(elf)\n+\n+ pgms = elf.getPheaders()\n+ for pgm in pgms:\n+ if pgm.p_type == Elf.PT_LOAD:\n+ if pgm.p_memsz == 0:\n+ continue\n+ logger.info('Loading: %s', pgm)\n+ bytez = elf.readAtOffset(pgm.p_offset, pgm.p_filesz)\n+ bytez += b'\\x00' * (pgm.p_memsz - pgm.p_filesz)\n+ pva = pgm.p_vaddr\n+ if addbase:\n+ pva += baseaddr\n+ memmaps.append((pva, pgm.p_flags & 0x7, fname, bytez, e_const.PAGE_SIZE))\n+ else:\n+ logger.info('Skipping: %s', pgm)\n+\n+ if len(pgms) == 0:\n+ secs = elf.getSections()\n+ # fall back to loading sections as best we can...\n+ vw.vprint('elf: no program headers found!')\n+\n+ maps = [ [s.sh_offset,s.sh_size] for s in secs if s.sh_offset and s.sh_size ]\n+ maps.sort()\n+\n+ merged = []\n+ for i in range(len(maps)):\n+\n+ if merged and maps[i][0] == (merged[-1][0] + merged[-1][1]):\n+ merged[-1][1] += maps[i][1]\n+ continue\n+\n+ merged.append( maps[i] )\n+\n+ baseaddr = 0x05000000\n+ for offset,size in merged:\n+ bytez = elf.readAtOffset(offset,size)\n+ memmaps.append((baseaddr + offset, 0x7, fname, bytez, None))\n+\n+ for sec in secs:\n+ if sec.sh_offset and sec.sh_size:\n+ sec.sh_addr = baseaddr + sec.sh_offset\n+\n+\n+\ndef makeStringTable(vw, va, maxva):\nwhile va < maxva:\n@@ -133,6 +214,17 @@ archcalls = {\n'thumb16': 'armcall',\n}\n+def getAddBaseAddr(elf, baseaddr=None):\n+ '''\n+ # NOTE: This is only for prelink'd so's and exe's. Make something for old style so.\n+ '''\n+ addbase = False\n+ if not elf.isPreLinked() and elf.isSharedObject():\n+ addbase = True\n+ if baseaddr is None:\n+ baseaddr = elf.getBaseAddress()\n+ return addbase, baseaddr\n+\ndef loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# analysis of discovered functions and data locations should be stored until the end of loading\ndata_ptrs = []\n@@ -172,12 +264,8 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nvw.setEndian(elf.getEndian())\n# Base addr is earliest section address rounded to pagesize\n- # NOTE: This is only for prelink'd so's and exe's. Make something for old style so.\n- addbase = False\n- if not elf.isPreLinked() and elf.isSharedObject():\n- addbase = True\n- if baseaddr is None:\n- baseaddr = elf.getBaseAddress()\n+ # Some ELF's require adding the baseaddr to most/all later addresses\n+ addbase, baseaddr = getAddBaseAddr(elf, baseaddr)\nelf.fd.seek(0)\nmd5hash = v_parsers.md5Bytes(byts)\n@@ -210,6 +298,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\npgms = elf.getPheaders()\nsecs = elf.getSections()\n+ '''\nfor pgm in pgms:\nif pgm.p_type == Elf.PT_LOAD:\nif pgm.p_memsz == 0:\n@@ -248,7 +337,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nfor sec in secs:\nif sec.sh_offset and sec.sh_size:\nsec.sh_addr = baseaddr + sec.sh_offset\n-\n+ '''\n# First add all section definitions so we have them\nfor sec in secs:\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/ihex.py",
"new_path": "vivisect/parsers/ihex.py",
"diff": "@@ -30,6 +30,7 @@ def parseFile(vw, filename, baseaddr=None):\nvw.setMeta('DefaultCall', archcalls.get(arch, 'unknown'))\n+ # figure out of there's an offset into the file we need to skip\noffset = vw.config.viv.parsers.ihex.offset\nif not offset:\noffset = 0\n@@ -65,3 +66,39 @@ def parseFile(vw, filename, baseaddr=None):\ndef parseMemory(vw, memobj, baseaddr):\nraise Exception('ihex loader cannot parse memory!')\n+\n+def getBaseAndSize(vw, filename, baseaddr=None):\n+ '''\n+ Returns the default baseaddr and memory size required to load the file\n+ '''\n+ savebase = baseaddr\n+\n+ # figure out of there's an offset into the file we need to skip\n+ offset = vw.config.viv.parsers.ihex.offset\n+ if not offset:\n+ offset = 0\n+ ihex = v_ihex.IHexFile()\n+ with open(filename, 'rb') as f:\n+ shdr = f.read(offset)\n+ sbytes = f.read()\n+\n+ ihex.vsParse(sbytes)\n+\n+ memmaps = ihex.geMemoryMaps()\n+ baseaddr = 0xffffffffffffffffffffffff\n+ topmem = 0\n+\n+ for mapva, mperms, mname, mbytes, malign in memmaps:\n+ if mapva < baseaddr:\n+ baseaddr = mapva\n+ endva = mapva + len(mbytes)\n+ if endva > topmem:\n+ topmem = endva\n+\n+ size = topmem - baseaddr\n+ if savebase:\n+ # if we provided a baseaddr, override what the file wants\n+ baseaddr = savebase\n+\n+ return baseaddr, size\n+\n"
},
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/macho/const.py",
"new_path": "vstruct/defs/macho/const.py",
"diff": "@@ -202,3 +202,14 @@ N_SECT = 0xe # defined in section number n_sect\nN_PBUD = 0xc # prebound undefined (defined in a dylib)\nN_INDR = 0xa # indirect\n+ENDIAN_LSB = 0\n+ENDIAN_MSB = 1\n+\n+hdr_info = {\n+ 'cefaedfe': (4, ENDIAN_LSB), # Mach-O Little Endian (32-bit)\n+ 'cffaedfe': (8, ENDIAN_LSB), # Mach-O Little Endian (64-bit)\n+ 'feedface': (4, ENDIAN_MSB), # Mach-O Big Endian (32-bit)\n+ 'feedfacf': (8, ENDIAN_MSB), # Mach-O Big Endian (64-bit)\n+ 'cafebabe': (0, -1), # Universal Binary Big Endian. These fat binaries are archives that can include binaries for multiple architectures, but typically contain PowerPC and Intel x86.\n+ 'bebafeca': (0, -1), # Universal Binary Big Endian. These fat binaries are archives that can include binaries for multiple architectures, but typically contain PowerPC and Intel x86.\n+}\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
lots of updates to support getBaseAddrAndSize() on each file format (except macho, which is a wip)
|
718,765 |
16.11.2021 10:55:18
| 18,000 |
f06aee9983a1bfd602cc3adb57dbabcba3834086
|
Delete Relocation
Add a delete relocation even handler
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -400,6 +400,22 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nself._fireEvent(VWE_ADDRELOC, (fname, offset, rtype, data))\nreturn self.getRelocation(va)\n+ def delRelocation(self, va, full=False):\n+ \"\"\"\n+ Delete a tracked relocation.\n+ \"\"\"\n+ mmap = self.getMemoryMap(va)\n+ if not mmap:\n+ logger.warning('delRelocation: No matching map found for %s', va)\n+ return None\n+\n+ mmva, mmsz, mmperm, fname = mmap # FIXME: getFileByVa does not obey file defs\n+ reloc = self.getRelocation(va)\n+ if not reloc:\n+ return None\n+ self._fireEvent(VWE_DELRELOC, (fname, va, reloc, full))\n+ return reloc\n+\ndef getRelocations(self):\n\"\"\"\nGet the current list of relocation entries.\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/base.py",
"new_path": "vivisect/base.py",
"diff": "@@ -261,6 +261,29 @@ class VivWorkspaceCore(viv_impapi.ImportApi):\nself._handleADDXREF((rva, ptr, REF_PTR, 0))\nself._handleADDLOCATION((rva, self.psize, LOC_POINTER, ptr))\n+ def _handleDELRELOC(self, einfo):\n+ fname, rva, rtyp, full = einfo\n+ imgbase = self.getFileMeta(fname, 'imagebase')\n+ ptroff = rva - imgbase\n+\n+ self.reloc_by_va.pop(rva, None)\n+ delidx = -1\n+\n+ for idx, (fn, off, typ, data) in enumerate(self.relocations):\n+ if fn == fname and off == ptroff and typ == rtyp:\n+ delidx = idx\n+ break\n+\n+ if delidx >= 0:\n+ self.relocations.pop(delidx)\n+\n+ if full:\n+ if rtyp == RTYPE_BASEPTR:\n+ ptr = imgbase + data\n+ ptr, reftype, rflags = self.arch.archModifyXrefAddr(ptr, None, None)\n+ self._handleDELXREF((rva, ptr, REF_PTR, 0))\n+ self._handleDELLOCATION((rva, self.psize, LOC_POINTER, ptr))\n+\ndef _handleADDMODULE(self, einfo):\nlogger.warning('DEPRECATED (ADDMODULE) ignored: %s', einfo)\n@@ -510,7 +533,7 @@ class VivWorkspaceCore(viv_impapi.ImportApi):\nself.ehand[VWE_ADDSEGMENT] = self._handleADDSEGMENT\nself.ehand[VWE_DELSEGMENT] = None\nself.ehand[VWE_ADDRELOC] = self._handleADDRELOC\n- self.ehand[VWE_DELRELOC] = None\n+ self.ehand[VWE_DELRELOC] = self._handleDELRELOC\nself.ehand[VWE_ADDMODULE] = self._handleADDMODULE\nself.ehand[VWE_DELMODULE] = self._handleDELMODULE\nself.ehand[VWE_ADDFMODULE] = self._handleADDFMODULE\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -12,6 +12,7 @@ import vivisect.const as v_const\nimport vivisect.tools.graphutil as v_t_graphutil\nimport vivisect.tests.helpers as helpers\n+import vivisect.tests.utils as v_t_utils\nlogger = logging.getLogger(__name__)\n@@ -25,12 +26,13 @@ def isint(x):\nreturn type(x) is int\n-class VivisectTest(unittest.TestCase):\n+class VivisectTest(v_t_utils.VivTest):\n@classmethod\ndef setUpClass(cls):\ncls.firefox_vw = helpers.getTestWorkspace('windows', 'amd64', 'firefox.exe')\ncls.chgrp_vw = helpers.getTestWorkspace('linux', 'i386', 'chgrp.llvm')\ncls.vdir_vw = helpers.getTestWorkspace('linux', 'i386', 'vdir.llvm')\n+ cls.chown_vw = helpers.getTestWorkspace('linux', 'amd64', 'chown')\nfor vw in cls.vdir_vw, cls.chgrp_vw, cls.firefox_vw:\noldcanv = vw.canvas\n@@ -1486,3 +1488,30 @@ class VivisectTest(unittest.TestCase):\n0x140049b40, 0x140049aa0, 0x1400497a0, 0x140049720, 0x140049820, 0x140049a20]\nself.assertEqual(thunks, set(impthunk))\n+\n+ def test_del_reloc(self):\n+ with self.snap(self.chown_vw) as vw:\n+ base = vw.getFileMeta('chown', 'imagebase')\n+\n+ va = 0x20f950\n+ rva = 0x20f950 + base\n+\n+ reloc = [rdat for rdat in vw.getRelocations() if rdat[1] == va]\n+ self.len(reloc, 1)\n+ rtyp = vw.getRelocation(rva)\n+ self.assertEqual(2, rtyp)\n+\n+ old = len(vw.getRelocations())\n+ self.assertEqual(2, vw.delRelocation(rva, full=True))\n+\n+ self.none(vw.getLocation(rva))\n+ self.len(vw.getXrefsFrom(rva), 0)\n+ self.len(vw.getXrefsTo(0x20028a0), 1)\n+ self.eq(2, self.chown_vw.getRelocation(rva))\n+ self.eq(1, old - len(vw.getRelocations()))\n+\n+ self.none(vw.delRelocation(0xabad1dea))\n+\n+ self.nn(self.chown_vw.getLocation(rva))\n+ self.len(self.chown_vw.getXrefsFrom(rva), 1)\n+ self.len(self.chown_vw.getXrefsTo(0x20028a0), 2)\n"
},
{
"change_type": "ADD",
"old_path": null,
"new_path": "vivisect/tests/utils.py",
"diff": "+import unittest\n+import contextlib\n+\n+import vivisect.cli as v_cli\n+\n+class VivTest(unittest.TestCase):\n+ '''\n+ Base class for vivisect unit tests so that we have a common place to throw certain common\n+ utilities.\n+ '''\n+\n+ def eq(self, x, y):\n+ '''\n+ Assert x is equal to y\n+ '''\n+ self.assertEqual(x, y)\n+\n+ def len(self, x, y):\n+ '''\n+ Assert that length of x is equal to y\n+ '''\n+ self.assertEqual(len(x), y)\n+\n+ def none(self, x):\n+ '''\n+ Assert x is none\n+ '''\n+ self.assertIsNone(x)\n+\n+ def nn(self, x):\n+ '''\n+ Assert x is not none\n+ '''\n+ self.assertIsNotNone(x)\n+\n+ @contextlib.contextmanager\n+ def snap(self, vw):\n+ '''\n+ Checkpoint a workspace. Yields a new workspace that can be editted\n+ as the test needs, and once the context handler ends, all changes will\n+ tossed\n+\n+ To be used with some caution, as it does create a duplicate workspace.\n+ '''\n+ safe = v_cli.VivCli()\n+ events = list(vw.exportWorkspace())\n+ safe.importWorkspace(events)\n+ yield safe\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Delete Relocation (#471)
Add a delete relocation even handler
|
718,770 |
16.11.2021 20:19:39
| 18,000 |
351dfe860273789efc7f6fffcaa1e33abc01e311
|
working on getMemBaseAndSize
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2759,7 +2759,7 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nmod = viv_parsers.getParserModule(fmtname)\n# get baseaddr and size, then make sure we have a good baseaddr\n- baseaddr, size = mod.getMemoryBaseAndSize(self, filename=filename, baseaddr=baseaddr)\n+ baseaddr, size = mod.getMemBaseAndSize(self, filename=filename, baseaddr=baseaddr)\nbaseaddr = self.findFreeMemoryBlock(size, baseaddr)\nlogger.debug(\"loading %r (size: 0x%x) at 0x%x\", filename, size, baseaddr)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/blob.py",
"new_path": "vivisect/parsers/blob.py",
"diff": "+import os\nimport envi\nimport vivisect.exc as v_exc\nimport vivisect.parsers as v_parsers\n@@ -78,7 +79,7 @@ def parseMemory(vw, memobj, baseaddr):\nvw.addMemoryMap(va, perms, fname, bytez)\nvw.setMeta('DefaultCall', archcalls.get(arch,'unknown'))\n-def getBaseAndSize(vw, filename, baseaddr=None):\n+def getMemBaseAndSize(vw, filename, baseaddr=None):\nif baseaddr is None:\nbaseaddr = vw.config.viv.parsers.blob.baseaddr\nsize = os.lstat(filename).st_size\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -43,7 +43,7 @@ def parseFd(vw, fd, filename=None, baseaddr=None):\ndef parseMemory(vw, memobj, baseaddr):\nraise Exception('FIXME implement parseMemory for elf!')\n-def getBaseAndSize(vw, filename, baseaddr=None):\n+def getMemBaseAndSize(vw, filename, baseaddr=None):\n'''\nReturns the default baseaddr and memory size required to load the file\n'''\n@@ -72,7 +72,7 @@ def getBaseAndSize(vw, filename, baseaddr=None):\n-def getMemoryMapInfo(elf):\n+def getMemoryMapInfo(elf, fname=None):\n'''\nGets the default baseaddr and memory map information\nAll the information necessary to add memory maps (or get overall size info)\n@@ -99,7 +99,7 @@ def getMemoryMapInfo(elf):\nif len(pgms) == 0:\nsecs = elf.getSections()\n# fall back to loading sections as best we can...\n- vw.vprint('elf: no program headers found!')\n+ logger.info('elf: no program headers found! (in %r)', fname)\nmaps = [ [s.sh_offset,s.sh_size] for s in secs if s.sh_offset and s.sh_size ]\nmaps.sort()\n@@ -122,6 +122,8 @@ def getMemoryMapInfo(elf):\nif sec.sh_offset and sec.sh_size:\nsec.sh_addr = baseaddr + sec.sh_offset\n+ return memmaps\n+\ndef makeStringTable(vw, va, maxva):\n@@ -295,49 +297,10 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nfor sec in elf.getSections():\nsecnames.append(sec.getName())\n- pgms = elf.getPheaders()\nsecs = elf.getSections()\n- '''\n- for pgm in pgms:\n- if pgm.p_type == Elf.PT_LOAD:\n- if pgm.p_memsz == 0:\n- continue\n- logger.info('Loading: %s', pgm)\n- bytez = elf.readAtOffset(pgm.p_offset, pgm.p_filesz)\n- bytez += b'\\x00' * (pgm.p_memsz - pgm.p_filesz)\n- pva = pgm.p_vaddr\n- if addbase:\n- pva += baseaddr\n- vw.addMemoryMap(pva, pgm.p_flags & 0x7, fname, bytez, align=e_const.PAGE_SIZE)\n- else:\n- logger.info('Skipping: %s', pgm)\n-\n- if len(pgms) == 0:\n- # fall back to loading sections as best we can...\n- vw.vprint('elf: no program headers found!')\n-\n- maps = [ [s.sh_offset,s.sh_size] for s in secs if s.sh_offset and s.sh_size ]\n- maps.sort()\n-\n- merged = []\n- for i in range(len(maps)):\n-\n- if merged and maps[i][0] == (merged[-1][0] + merged[-1][1]):\n- merged[-1][1] += maps[i][1]\n- continue\n-\n- merged.append( maps[i] )\n-\n- baseaddr = 0x05000000\n- for offset,size in merged:\n- bytez = elf.readAtOffset(offset,size)\n- vw.addMemoryMap(baseaddr + offset, 0x7, fname, bytez)\n-\n- for sec in secs:\n- if sec.sh_offset and sec.sh_size:\n- sec.sh_addr = baseaddr + sec.sh_offset\n- '''\n+ for mmapva, mmperms, mfname, mbytez, malign in getMemoryMapInfo(elf, fname):\n+ vw.addMemoryMap(mmapva, mmperms, mfname, mbytez, malign)\n# First add all section definitions so we have them\nfor sec in secs:\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/ihex.py",
"new_path": "vivisect/parsers/ihex.py",
"diff": "@@ -67,7 +67,7 @@ def parseFile(vw, filename, baseaddr=None):\ndef parseMemory(vw, memobj, baseaddr):\nraise Exception('ihex loader cannot parse memory!')\n-def getBaseAndSize(vw, filename, baseaddr=None):\n+def getMemBaseAndSize(vw, filename, baseaddr=None):\n'''\nReturns the default baseaddr and memory size required to load the file\n'''\n@@ -84,11 +84,11 @@ def getBaseAndSize(vw, filename, baseaddr=None):\nihex.vsParse(sbytes)\n- memmaps = ihex.geMemoryMaps()\n+ memmaps = ihex.getMemoryMaps()\nbaseaddr = 0xffffffffffffffffffffffff\ntopmem = 0\n- for mapva, mperms, mname, mbytes, malign in memmaps:\n+ for mapva, mperms, mname, mbytes in memmaps:\nif mapva < baseaddr:\nbaseaddr = mapva\nendva = mapva + len(mbytes)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/srec.py",
"new_path": "vivisect/parsers/srec.py",
"diff": "@@ -64,3 +64,39 @@ def parseFile(vw, filename, baseaddr=None):\ndef parseMemory(vw, memobj, baseaddr):\nraise Exception('srec loader cannot parse memory!')\n+\n+def getMemBaseAndSize(vw, filename, baseaddr=None):\n+ '''\n+ Returns the default baseaddr and memory size required to load the file\n+ '''\n+ savebase = baseaddr\n+\n+ # figure out of there's an offset into the file we need to skip\n+ offset = vw.config.viv.parsers.srec.offset\n+ if not offset:\n+ offset = 0\n+ srec = v_srec.SRecFile()\n+ with open(filename, 'rb') as f:\n+ shdr = f.read(offset)\n+ sbytes = f.read()\n+\n+ srec.vsParse(sbytes)\n+\n+ memmaps = srec.getMemoryMaps()\n+ baseaddr = 0xffffffffffffffffffffffff\n+ topmem = 0\n+\n+ for mapva, mperms, mname, mbytes in memmaps:\n+ if mapva < baseaddr:\n+ baseaddr = mapva\n+ endva = mapva + len(mbytes)\n+ if endva > topmem:\n+ topmem = endva\n+\n+ size = topmem - baseaddr\n+ if savebase:\n+ # if we provided a baseaddr, override what the file wants\n+ baseaddr = savebase\n+\n+ return baseaddr, size\n+\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
working on getMemBaseAndSize
|
718,770 |
18.11.2021 23:38:07
| 18,000 |
dfb3d8554e2291dd45f7d3c35655e689bbba9625
|
bugfix: treating baseaddr as if it is only as provided by DLL. separated out "imagebase" and "baseaddr".
making progress on loading. not done yet... but the baseaddr bug helps.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2794,6 +2794,59 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\n# Snapin our analysis modules\nself._snapInAnalysisModules()\n+ def writeMemory(self, va, bytez):\n+ '''\n+ Override writeMemory to hook into the Event subsystem.\n+ '''\n+ fname = self.getFileByVa(va)\n+ fileva = self.getFileMeta(fname, 'imagebase')\n+ off = va - fileva\n+ self._fireEvent(VWE_WRITEMEM, (fname, off, bytez, self._supervisor))\n+\n+ def connectImportsWithExports(self):\n+ \"\"\"\n+ Look for any \"imported\" symbols that are satisfied by current exports.\n+ Wire up the connection.\n+\n+ Currently this is simply a pointer write at the location of the Import.\n+ If this behavior is ever insufficient, we'll want to track the special\n+ nature through the Export/Import events.\n+ \"\"\"\n+ # store old setting and set _supervisor mode (so we can write wherever\n+ # we want, regardless of permissions)\n+ oldsup = self._supervisor\n+ self._supervisor = True\n+\n+ for iva, isz, itype, isym in self.getImports():\n+ impfname, impsym = isym.split('.', 1)\n+ for eva, num, esym, efname in self.getExports():\n+ if impsym != esym:\n+ continue\n+\n+ if impfname not in (efname, '*'):\n+ continue\n+\n+ # file and symbol name match. apply the magic.\n+ # do we ever *not* write in the full address at the import site?\n+ self.writeMemoryPtr(iva, eva)\n+\n+ # check if any xrefs to the import are branches and make code-xrefs for them\n+ for xrfr, xrto, xrt, xrflags in self.getXrefsTo(iva):\n+ loc = self.getLocation(xrfr)\n+ if not loc:\n+ continue\n+\n+ lva, lsz, ltype, ltinfo = loc\n+ if ltype != LOC_OP:\n+ logger.warning(\"XREF not from an Opcode: 0x%x -> 0x%x (%r)\", lva, eva, loc)\n+ continue\n+\n+ op = self.parseOpcode(lva)\n+ self.addXref(lva, eva, REF_CODE)\n+\n+ # restore previous supervisor mode\n+ self._supervisor = oldsup\n+\ndef getFiles(self):\n\"\"\"\nReturn the current list of file objects in this\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/base.py",
"new_path": "vivisect/base.py",
"diff": "@@ -526,6 +526,18 @@ class VivWorkspaceCore(viv_impapi.ImportApi):\n'''\npass\n+ def _handleWRITEMEM(self, einfo):\n+ '''\n+ Handle permanent writes to a memory map after initialization\n+ (fname, off, bytez, supv) where supv is supervisor mode...\n+ '''\n+ fname, off, bytez, supv = einfo\n+ imgbase = self.getFileMeta(fname, 'imagebase')\n+ savesupv = self._supervisor\n+ self._supervisor = True\n+ e_mem.MemoryObject.writeMemory(self, imgbase + off, bytez)\n+ self._supervisor = savesupv\n+\ndef _initEventHandlers(self):\nself.ehand = [None for x in range(VWE_MAX)]\nself.ehand[VWE_ADDLOCATION] = self._handleADDLOCATION\n@@ -569,6 +581,7 @@ class VivWorkspaceCore(viv_impapi.ImportApi):\nself.ehand[VWE_CHAT] = self._handleCHAT\nself.ehand[VWE_SYMHINT] = self._handleSYMHINT\nself.ehand[VWE_AUTOANALFIN] = self._handleAUTOANALFIN\n+ self.ehand[VWE_WRITEMEM] = self._handleWRITEMEM\nself.thand = [None for x in range(VTE_MAX)]\nself.thand[VTE_IAMLEADER] = self._handleIAMLEADER\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/const.py",
"new_path": "vivisect/const.py",
"diff": "@@ -71,7 +71,9 @@ VWE_CHAT = 40 # (username, message)\nVWE_SYMHINT = 41 # (va, idx, hint)\nVWE_AUTOANALFIN = 42 # (starttime, endtime)\n-VWE_MAX = 43\n+VWE_WRITEMEM = 43 # (mapva, offset, bytes)\n+\n+VWE_MAX = 44\n# Constants for vivisect \"transient\" events which flow through\n# the event subsystem but are not recorded to the workspace.\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/pe.py",
"new_path": "vivisect/parsers/pe.py",
"diff": "@@ -13,6 +13,7 @@ import vtrace # needed only for setting the logging level\nimport vtrace.platforms.win32 as vt_win32\nimport envi.exc as e_exc\n+import envi.bits as e_bits\nimport envi.const as e_const\nimport envi.symstore.symcache as e_symcache\n@@ -76,8 +77,8 @@ archcalls = {\n# map PE relocation types to vivisect types where possible\nrelmap = {\n- PE.IMAGE_REL_BASED_HIGHLOW: vivisect.RTYPE_BASEOFF,\n- PE.IMAGE_REL_BASED_DIR64: vivisect.RTYPE_BASEOFF,\n+ PE.IMAGE_REL_BASED_HIGHLOW: (vivisect.RTYPE_BASEOFF, 4),\n+ PE.IMAGE_REL_BASED_DIR64: (vivisect.RTYPE_BASEOFF, 8),\n}\n@@ -105,18 +106,24 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\nvw.setMeta('Platform', platform)\nvw.setMeta('DefaultCall', archcalls.get(arch, 'unknown'))\n+ logger.info(\"PE loader: Arch: %r\\tFormat: pe\\tPlatform: %r\\tFilename: %r\\tBaseAddr: 0x%x\", \\\n+ arch, platform, filename, baseaddr)\n# Set ourselves up for extended windows binary analysis\n+ imagebase = pe.IMAGE_NT_HEADERS.OptionalHeader.ImageBase\nif baseaddr is None:\n- baseaddr = pe.IMAGE_NT_HEADERS.OptionalHeader.ImageBase\n- entry = pe.IMAGE_NT_HEADERS.OptionalHeader.AddressOfEntryPoint + baseaddr\n- entryrva = entry - baseaddr\n+ baseaddr = imagebase\n+\n+ entryrva = pe.IMAGE_NT_HEADERS.OptionalHeader.AddressOfEntryPoint\n+ entry = entryrva + baseaddr\ncodebase = pe.IMAGE_NT_HEADERS.OptionalHeader.BaseOfCode\ncodesize = pe.IMAGE_NT_HEADERS.OptionalHeader.SizeOfCode\ncodervamax = codebase+codesize\n+ logger.info(\"PE Imagebase: 0x%x\\tentry: 0x%x\\tcodebase: 0x%x\\tcodesize: 0x%x\", \\\n+ imagebase, entry, codebase, codesize)\n# grab the file bytes for hashing\npe.fd.seek(0)\nfhash = v_parsers.md5Bytes(byts)\n@@ -131,6 +138,9 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\nif fvivname is None:\nfvivname = fhash\n+ logger.info(\"PE dllname: %r\\tfvivname: %r\\tmd5: %r\\tsha256: %r\", \\\n+ dllname, fvivname, fhash, sha256)\n+\n# create the file and store md5 and sha256 hashes\nfname = vw.addFile(fvivname.lower(), baseaddr, fhash)\nvw.setFileMeta(fname, 'sha256', sha256)\n@@ -350,15 +360,18 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\nfor rva, rtype in pe.getRelocations():\n# map PE reloc to VIV reloc ( or dont... )\n- vtype = relmap.get(rtype)\n- if vtype is None:\n+ vtypedata = relmap.get(rtype)\n+ if vtypedata is None:\nlogger.info('Skipping PE Relocation type: %d at %d (no handler)', rtype, rva)\ncontinue\n+ vtype, vtsize = vtypedata\n+\ntry:\n- mapoffset = vw.readMemoryPtr(rva + baseaddr) - baseaddr\n- except:\n- # the target adderss of the relocation is not accessible.\n+ vtfmt = e_bits.getFormat(vtsize) # for PowerPC, will need to get Endianness\n+ mapoffset = vw.readMemoryFormat(rva + baseaddr, vtfmt)[0] - imagebase\n+ except Exception as e:\n+ # the target address of the relocation is not accessible.\n# for example, it's not mapped, or split across sections, etc.\n# technically, the PE is corrupt.\n# by continuing on here, we are a bit more robust (but maybe incorrect)\n@@ -366,9 +379,10 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\n#\n# discussed in:\n# https://github.com/vivisect/vivisect/issues/346\n- logger.warning('Skipping invalid PE relocation: %d', rva)\n+ logger.warning('Skipping invalid PE relocation: %d (%r)', rva, e)\ncontinue\nelse:\n+ logger.info('PE relocation: 0x%x -> %r+0x%x', baseaddr+rva, fname, mapoffset)\nvw.addRelocation(rva + baseaddr, vtype, mapoffset)\nfor rva, lname, iname in pe.getImports():\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
bugfix: treating baseaddr as if it is only as provided by DLL. separated out "imagebase" and "baseaddr".
making progress on loading. not done yet... but the baseaddr bug helps.
|
718,770 |
23.11.2021 14:54:47
| 18,000 |
0e088d7c32bd9aec20bba94b8f13d24a1107cb2c
|
feature: -o to select the output filename
bugfix: libc_start_main() errored out with a string mismatch
bugfix: move "connectExportsToImports()" to it's own "linker" analysis module
bugfix: ELF mobile loading ability!!!!!!!!!!!! <AGCK!>
feature: new VaSet "ResolvedImports" for imports that have been resolved and turned into Pointers.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -176,6 +176,8 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nself.addVaSet('CodeFragments', (('va', VASET_ADDRESS), ('calls_from', VASET_COMPLEX)))\nself.addVaSet('EmucodeFunctions', (('va', VASET_ADDRESS),))\nself.addVaSet('FuncWrappers', (('va', VASET_ADDRESS), ('wrapped_va', VASET_ADDRESS),))\n+ self.addVaSet('ResolvedImports', (('va',VASET_ADDRESS), ('symbol', VASET_STRING), \\\n+ ('resolved address', VASET_ADDRESS)))\ndef vprint(self, msg):\nlogger.info(msg)\n@@ -2758,13 +2760,19 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nreturn self.normFileName(filename)\nmod = viv_parsers.getParserModule(fmtname)\n+\n# get baseaddr and size, then make sure we have a good baseaddr\nbaseaddr, size = mod.getMemBaseAndSize(self, filename=filename, baseaddr=baseaddr)\n+ logger.debug('initial baseva: 0x%x size: 0x%x', baseaddr, size)\n+ if baseaddr == 0:\n+ baseaddr = 0x300000\n+\nbaseaddr = self.findFreeMemoryBlock(size, baseaddr)\nlogger.debug(\"loading %r (size: 0x%x) at 0x%x\", filename, size, baseaddr)\nfname = mod.parseFile(self, filename=filename, baseaddr=baseaddr)\n+ if not self.getMeta('StorageName'):\nself.initMeta(\"StorageName\", filename+\".viv\")\n# Snapin our analysis modules\n@@ -2812,6 +2820,7 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nIf this behavior is ever insufficient, we'll want to track the special\nnature through the Export/Import events.\n\"\"\"\n+ logger.info('linking Imports with Exports')\n# store old setting and set _supervisor mode (so we can write wherever\n# we want, regardless of permissions)\noldsup = self._supervisor\n@@ -2828,8 +2837,16 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\n# file and symbol name match. apply the magic.\n# do we ever *not* write in the full address at the import site?\n+ logger.debug(\"connecting Import 0x%x -> Export 0x%x (%r)\", iva, eva, isym)\nself.writeMemoryPtr(iva, eva)\n+ # remove the LOC_IMPORT and make it a Pointer instead\n+ self.delLocation(iva)\n+ self.makePointer(iva, follow=False) # don't follow, it'll be analyzed later?\n+\n+ # store the former Import in a VaSet\n+ self.setVaSetRow('ResolvedImports', (iva, isym, eva))\n+\n# check if any xrefs to the import are branches and make code-xrefs for them\nfor xrfr, xrto, xrt, xrflags in self.getXrefsTo(iva):\nloc = self.getLocation(xrfr)\n@@ -2843,6 +2860,7 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nop = self.parseOpcode(lva)\nself.addXref(lva, eva, REF_CODE)\n+ logger.debug(\"addXref(0x%x -> 0x%x)\", lva, eva)\n# restore previous supervisor mode\nself._supervisor = oldsup\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/__init__.py",
"new_path": "vivisect/analysis/__init__.py",
"diff": "@@ -20,6 +20,7 @@ def addAnalysisModules(vw):\nif fmt == 'pe':\n+ vw.addAnalysisModule(\"vivisect.analysis.generic.linker\")\nvw.addAnalysisModule(\"vivisect.analysis.generic.entrypoints\")\nvw.addAnalysisModule(\"vivisect.analysis.pe\")\n@@ -90,6 +91,7 @@ def addAnalysisModules(vw):\n# elfplt wants to be run before generic.entrypoints.\nvw.addAnalysisModule(\"vivisect.analysis.elf.elfplt\")\n+ vw.addAnalysisModule(\"vivisect.analysis.generic.linker\")\nvw.addAnalysisModule(\"vivisect.analysis.generic.entrypoints\")\nvw.addAnalysisModule(\"vivisect.analysis.elf\")\n@@ -143,6 +145,7 @@ def addAnalysisModules(vw):\nelif fmt == 'macho': # MACH-O ###################################################\n+ vw.addAnalysisModule(\"vivisect.analysis.generic.linker\")\nvw.addAnalysisModule(\"vivisect.analysis.generic.entrypoints\")\nif arch == 'i386':\nviv_analysis_i386.addEntrySigs(vw)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/elfplt.py",
"new_path": "vivisect/analysis/elf/elfplt.py",
"diff": "@@ -104,9 +104,9 @@ def getGOTByFilename(vw, filename):\nif FGOT is not None:\n# be sure to add the imgbase to FGOT if required\nif vw.getFileMeta(filename, 'addbase'):\n- imgbase = vw.getFileMeta(filename, 'imagebase')\n- logger.debug('Adding Imagebase: 0x%x', imgbase)\n+ imgbase = vw.getFileMeta(filename, 'baseoff')\nFGOT += imgbase\n+ logger.debug('Adding Imagebase(%r) for Dynamics GOT: 0x%x (0x%x)', filename, imgbase, FGOT)\nif FGOT != gotva and None not in (FGOT, gotva):\nlogger.warning(\"Dynamics and Sections have different GOT entries: S:0x%x D:0x%x. using Dynamics\", gotva, FGOT)\n@@ -137,9 +137,9 @@ def getGOTs(vw):\nif FGOT is not None:\n# be sure to add the imgbase to FGOT if required\nif vw.getFileMeta(filename, 'addbase'):\n- imgbase = vw.getFileMeta(filename, 'imagebase')\n- #logger.debug('Adding Imagebase: 0x%x', imgbase)\n+ imgbase = vw.getFileMeta(filename, 'baseoff')\nFGOT += imgbase\n+ logger.debug('Adding Imagebase(%r) for Dynamics GOT: 0x%x (0x%x)', filename, imgbase, FGOT)\nflist = out[filename]\n@@ -170,19 +170,25 @@ def getPLTs(vw):\nfor fname in vw.getFiles():\nfdyns = vw.getFileMeta(fname, 'ELF_DYNAMICS')\nif fdyns is not None:\n- FGOT = fdyns.get('DT_JMPREL')\n- FGOTSZ = fdyns.get('DT_PLTRELSZ')\n+ FPLT = fdyns.get('DT_JMPREL')\n+ FPLTSZ = fdyns.get('DT_PLTRELSZ')\n+\n+ # if we don't have FPLT or FPLTSZ, skip this\n+ if None in (FPLT, FPLTSZ):\n+ continue\n+\nif vw.getFileMeta(fname, 'addbase'):\n- imgbase = vw.getFileMeta(fname, 'imagebase')\n- logger.debug('Adding Imagebase: 0x%x', imgbase)\n- FGOT += imgbase\n+ imgbase = vw.getFileMeta(fname, 'baseoff')\n+ FPLT += imgbase\n+ logger.debug('Adding Imagebase(%r): 0x%x (0x%x)', fname, imgbase, FPLT)\nnewish = True\nfor pltva, pltsize in plts:\n- if FGOT == pltva:\n+ if FPLT == pltva:\nnewish = False\n- if newish and FGOT and FGOTSZ:\n- plts.append((FGOT, FGOTSZ))\n+\n+ if newish:\n+ plts.append((FPLT, FPLTSZ))\nreturn plts\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/libc_start_main.py",
"new_path": "vivisect/analysis/elf/libc_start_main.py",
"diff": "@@ -87,7 +87,7 @@ class AnalysisMonitor(viv_imp_monitor.AnalysisMonitor):\nif op.iflags & envi.IF_CALL:\n# it's a call, get the target\nbranches = [br for br in op.getBranches(emu) if not (br[1] & envi.BR_FALL)]\n- logger.debug('libc_start_main: 0x%x\\ttgts: %r', self.startmain, branches)\n+ logger.debug('libc_start_main: %r\\ttgts: %r', [hex(x) for x in self.startmain], branches)\n# check if it matches what we believe to be __libc_start_main\nfor branch in branches:\n"
},
{
"change_type": "ADD",
"old_path": null,
"new_path": "vivisect/analysis/generic/linker.py",
"diff": "+'''\n+Connect any Exports we have to Imports we may also have\n+(most useful for multiple file workspaces)\n+'''\n+def analyze(vw):\n+ #vw.connectImportsWithExports()\n+ pass\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -72,14 +72,14 @@ def getMemBaseAndSize(vw, filename, baseaddr=None):\n-def getMemoryMapInfo(elf, fname=None):\n+def getMemoryMapInfo(elf, fname=None, baseaddr=None):\n'''\nGets the default baseaddr and memory map information\nAll the information necessary to add memory maps (or get overall size info)\n'''\nmemmaps = []\n- addbase, baseaddr = getAddBaseAddr(elf)\n+ addbase, baseoff, baseaddr = getAddBaseAddr(elf, baseaddr)\npgms = elf.getPheaders()\nfor pgm in pgms:\n@@ -90,8 +90,7 @@ def getMemoryMapInfo(elf, fname=None):\nbytez = elf.readAtOffset(pgm.p_offset, pgm.p_filesz)\nbytez += b'\\x00' * (pgm.p_memsz - pgm.p_filesz)\npva = pgm.p_vaddr\n- if addbase:\n- pva += baseaddr\n+ pva += baseoff\nmemmaps.append((pva, pgm.p_flags & 0x7, fname, bytez, e_const.PAGE_SIZE))\nelse:\nlogger.info('Skipping: %s', pgm)\n@@ -151,7 +150,8 @@ def makeSymbolTable(vw, va, maxva):\nva += len(s)\nreturn ret\n-def makeDynamicTable(vw, va, maxva):\n+def makeDynamicTable(vw, va, maxva, baseoff=0):\n+ logger.debug(\"0x%x -> 0x%x\", va, maxva)\nret = []\nsname = 'elf.Elf%dDynamic' % (vw.getPointerSize() * 8)\nwhile va < maxva:\n@@ -165,7 +165,7 @@ def makeDynamicTable(vw, va, maxva):\nbreak\nreturn ret\n-def makeRelocTable(vw, va, maxva, addbase, baseaddr, addend=False):\n+def makeRelocTable(vw, va, maxva, baseoff, addend=False):\nif addend:\nsname = 'elf.Elf%dReloca' % (vw.getPointerSize() * 8)\nelse:\n@@ -177,25 +177,23 @@ def makeRelocTable(vw, va, maxva, addbase, baseaddr, addend=False):\nvw.setComment(va, tname)\nva += len(s)\n-def makeFunctionTable(elf, vw, tbladdr, size, tblname, funcs, ptrs, baseaddr=0, addbase=False):\n- logger.debug('makeFunctionTable(tbladdr=0x%x, size=0x%x, tblname=%r, baseaddr=0x%x)', tbladdr, size, tblname, baseaddr)\n+def makeFunctionTable(elf, vw, tbladdr, size, tblname, funcs, ptrs, baseoff=0):\n+ logger.debug('makeFunctionTable(tbladdr=0x%x, size=0x%x, tblname=%r, baseoff=0x%x)', tbladdr, size, tblname, baseoff)\npsize = vw.getPointerSize()\nfmtgrps = e_bits.fmt_chars[vw.getEndian()]\npfmt = fmtgrps[psize]\nsecbytes = elf.readAtRva(tbladdr, size)\n- if addbase:\n- tbladdr += baseaddr\n+ tbladdr += baseoff\nptr_count = 0\nfor off in range(0, size, psize):\naddr, = struct.unpack_from(pfmt, secbytes, off)\n- if addbase:\n- addr += baseaddr\n+ addr += baseoff\nnstub = tblname + \"_%d\"\npname = nstub % ptr_count\n- vw.makeName(addr, pname, filelocal=True)\n+ vw.makeName(addr, pname, filelocal=True, makeuniq=True)\nptrs.append((tbladdr + off, addr, pname))\nfuncs.append((pname, addr))\nptr_count += 1\n@@ -220,15 +218,25 @@ def getAddBaseAddr(elf, baseaddr=None):\n'''\n# NOTE: This is only for prelink'd so's and exe's. Make something for old style so.\n'''\n- addbase = False\n+ baseoff = 0\n+ logger.debug(\"baseaddr: %r\\tbaseoff: 0x%x\", baseaddr, baseoff)\n+ elfbaseaddr = elf.getBaseAddress()\n+ if baseaddr is None:\n+ baseaddr = elfbaseaddr\n+\nif not elf.isPreLinked() and elf.isSharedObject():\naddbase = True\n- if baseaddr is None:\n- baseaddr = elf.getBaseAddress()\n- return addbase, baseaddr\n+ baseoff = baseaddr\n+ else:\n+ addbase = False\n+ baseoff = baseaddr - elfbaseaddr\n+\n+ logger.debug(\"baseaddr: %r\\tbaseoff: 0x%x\", baseaddr, baseoff)\n+ return addbase, baseoff, baseaddr\ndef loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# analysis of discovered functions and data locations should be stored until the end of loading\n+ logger.info(\"loadElfIntoWorkspace(filename=%r, baseaddr: %r\", filename, baseaddr)\ndata_ptrs = []\nnew_pointers = []\nnew_functions = []\n@@ -267,7 +275,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# Base addr is earliest section address rounded to pagesize\n# Some ELF's require adding the baseaddr to most/all later addresses\n- addbase, baseaddr = getAddBaseAddr(elf, baseaddr)\n+ addbase, baseoff, baseaddr = getAddBaseAddr(elf, baseaddr)\nelf.fd.seek(0)\nmd5hash = v_parsers.md5Bytes(byts)\n@@ -299,7 +307,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nsecs = elf.getSections()\n- for mmapva, mmperms, mfname, mbytez, malign in getMemoryMapInfo(elf, fname):\n+ for mmapva, mmperms, mfname, mbytez, malign in getMemoryMapInfo(elf, fname, baseaddr):\nvw.addMemoryMap(mmapva, mmperms, mfname, mbytez, malign)\n# First add all section definitions so we have them\n@@ -310,8 +318,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\ncontinue # Skip non-memory mapped sections\nsva = sec.sh_addr\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\nvw.addSegment(sva, size, sname, fname)\n@@ -326,8 +333,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\ncontinue\nsva = phdr.p_vaddr\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\nvw.addSegment(sva, phdr.p_memsz, 'PHDR%d' % pcount, fname)\npcount += 1\n@@ -335,54 +341,49 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# load information from dynamics:\nf_init = elf.dyns.get(Elf.DT_INIT)\nif f_init is not None:\n- if addbase:\n- f_init += baseaddr\n- vw.makeName(f_init, \"init_function\", filelocal=True)\n+ f_init += baseoff\n+ vw.makeName(f_init, \"init_function\", filelocal=True, makeuniq=True)\nvw.addEntryPoint(f_init)\nf_fini = elf.dyns.get(Elf.DT_FINI)\nif f_fini is not None:\n- if addbase:\n- f_fini += baseaddr\n- vw.makeName(f_fini, \"fini_function\", filelocal=True)\n+ f_fini += baseoff\n+ vw.makeName(f_fini, \"fini_function\", filelocal=True, makeuniq=True)\nvw.addEntryPoint(f_fini)\nf_inita = elf.dyns.get(Elf.DT_INIT_ARRAY)\nif f_inita is not None:\nf_initasz = elf.dyns.get(Elf.DT_INIT_ARRAYSZ)\n- makeFunctionTable(elf, vw, f_inita, f_initasz, 'init_array', new_functions, new_pointers, baseaddr, addbase)\n+ makeFunctionTable(elf, vw, f_inita, f_initasz, 'init_array', new_functions, new_pointers, baseoff)\nf_finia = elf.dyns.get(Elf.DT_FINI_ARRAY)\nif f_finia is not None:\nf_finiasz = elf.dyns.get(Elf.DT_FINI_ARRAYSZ)\n- makeFunctionTable(elf, vw, f_finia, f_finiasz, 'fini_array', new_functions, new_pointers, baseaddr, addbase)\n+ makeFunctionTable(elf, vw, f_finia, f_finiasz, 'fini_array', new_functions, new_pointers, baseoff)\nf_preinita = elf.dyns.get(Elf.DT_PREINIT_ARRAY)\nif f_preinita is not None:\nf_preinitasz = elf.dyns.get(Elf.DT_PREINIT_ARRAY)\n- makeFunctionTable(elf, vw, f_preinita, f_preinitasz, 'preinit_array', new_functions, new_pointers, baseaddr, addbase)\n+ makeFunctionTable(elf, vw, f_preinita, f_preinitasz, 'preinit_array', new_functions, new_pointers, baseoff)\n# dynamic table\nphdr = elf.getDynPHdr() # file offset?\nif phdr is not None:\nsva, size = phdr.p_vaddr, phdr.p_memsz\n- if addbase:\n- sva += baseaddr # getDynInfo returns (offset, filesz)\n- makeDynamicTable(vw, sva, sva+size)\n+ sva += baseoff # getDynInfo returns (offset, filesz)\n+ makeDynamicTable(vw, sva, sva+size, baseoff)\n# if there's no Dynamics PHDR, don't bother trying to parse it from Sections. It doesn't exist.\n# dynstr table\nsva, size = elf.getDynStrTabInfo()\nif sva is not None:\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\nmakeStringTable(vw, sva, sva+size)\n# dynsyms table\nsva, symsz, size = elf.getDynSymTabInfo()\nif sva is not None:\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\n[s for s in makeSymbolTable(vw, sva, sva+size)]\n# Now trigger section specific analysis\n@@ -394,8 +395,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\ncontinue # Skip non-memory mapped sections\nsva = sec.sh_addr\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\n# if we've already defined a location at this address, skip it. (eg. DYNAMICS)\nif vw.getLocation(sva) == sva:\n@@ -411,7 +411,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nnew_functions.append(init_tup)\nelif sname == \".init_array\":\n- makeFunctionTable(elf, vw, sec.sh_addr, size, 'init_function', new_functions, new_pointers, baseaddr, addbase)\n+ makeFunctionTable(elf, vw, sec.sh_addr, size, 'init_function', new_functions, new_pointers, baseoff)\nelif sname == \".fini\":\nfini_tup = (\"fini_function\", sva)\n@@ -420,10 +420,10 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nnew_functions.append(fini_tup)\nelif sname == \".fini_array\":\n- makeFunctionTable(elf, vw, sec.sh_addr, size, 'fini_function', new_functions, new_pointers, baseaddr, addbase)\n+ makeFunctionTable(elf, vw, sec.sh_addr, size, 'fini_function', new_functions, new_pointers, baseoff)\nelif sname == \".dynamic\": # Imports\n- makeDynamicTable(vw, sva, sva+size)\n+ makeDynamicTable(vw, sva, sva+size, baseoff)\nelif sname == \".dynstr\": # String table for dynamics\nmakeStringTable(vw, sva, sva+size)\n@@ -439,10 +439,10 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nmakeSymbolTable(vw, sva, sva+size)\nelif sec.sh_type == Elf.SHT_REL:\n- makeRelocTable(vw, sva, sva+size, addbase, baseaddr)\n+ makeRelocTable(vw, sva, sva+size, baseoff)\nelif sec.sh_type == Elf.SHT_RELA:\n- makeRelocTable(vw, sva, sva+size, addbase, baseaddr, addend=True)\n+ makeRelocTable(vw, sva, sva+size, baseoff, addend=True)\nif sec.sh_flags & Elf.SHF_STRINGS:\nmakeStringTable(vw, sva, sva+size)\n@@ -463,11 +463,12 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# but isn't currently accessible from the gui\nvw.setFileMeta(fname, 'ELF_DYNAMICS', elfmeta)\nvw.setFileMeta(fname, 'addbase', addbase)\n+ vw.setFileMeta(fname, 'baseoff', baseoff)\n# applyRelocs is specifically prior to \"process Dynamic Symbols\" because Dynamics-only symbols\n# (ie. not using Section Headers) may not get all the symbols. Some ELF's simply list too\n# small a space using SYMTAB and SYMTABSZ\n- postfix = applyRelocs(elf, vw, addbase, baseaddr)\n+ postfix = applyRelocs(elf, vw, addbase, baseoff)\n# process Dynamic Symbols - this must happen *after* relocations, which can expand the size of this\nfor s in elf.getDynSyms():\n@@ -476,8 +477,8 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nif sva == 0:\ncontinue\n- if addbase:\n- sva += baseaddr\n+\n+ sva += baseoff\nif sva == 0:\ncontinue\n@@ -504,8 +505,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# So aparently Elf64 binaries on amd64 use HIOS and then\n# s.st_other cause that's what all the kewl kids are doing...\nsva = s.st_other\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\nif vw.isValidPointer(sva):\ntry:\nnew_functions.append((\"DynSym: STT_HIOS\", sva))\n@@ -516,8 +516,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nelif stype == Elf.STT_MDPROC: # there's only one that isn't HI or LO...\nsva = s.st_other\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\nif vw.isValidPointer(sva):\ntry:\nvw.addExport(sva, EXP_DATA, dmglname, fname, makeuniq=True)\n@@ -530,10 +529,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nif dmglname in postfix:\nfor rlva, addend in postfix[dmglname]:\n- if addbase:\n- vw.addRelocation(rlva, RTYPE_BASEPTR, sva + addend - baseaddr)\n- else:\n- vw.addRelocation(rlva, RTYPE_BASEPTR, sva + addend)\n+ vw.addRelocation(rlva, RTYPE_BASEPTR, sva + addend - baseoff)\nvw.addVaSet(\"FileSymbols\", ((\"Name\", VASET_STRING), (\"va\", VASET_ADDRESS)))\nvw.addVaSet(\"WeakSymbols\", ((\"Name\", VASET_STRING), (\"va\", VASET_ADDRESS)))\n@@ -559,8 +555,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# mapping symbol\nif arch in ('arm', 'thumb', 'thumb16'):\nsymname = s.getName()\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\nif symname == '$a':\n# ARM code\nlogger.info('mapping (NOTYPE) ARM symbol: 0x%x: %r', sva, dmglname)\n@@ -577,8 +572,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\ndata_ptrs.append(sva)\nelif s.getInfoType() == Elf.STT_OBJECT:\nsymname = s.getName()\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\nif symname:\nvw.makeName(sva, symname, filelocal=True, makeuniq=True)\nvalu = vw.readMemoryPtr(sva)\n@@ -623,8 +617,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# TODO: make use of _ZTSN (typeinfo) and _ZTVN (vtable) entries if name demangles cleanly\n- if addbase:\n- sva += baseaddr\n+ sva += baseoff\nif vw.isValidPointer(sva) and len(dmglname):\ntry:\nif s.getInfoBind() == Elf.STB_WEAK:\n@@ -645,10 +638,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nif s.st_info == Elf.STT_FUNC:\nnew_functions.append((\"STT_FUNC\", sva))\n- if addbase:\n- eentry = baseaddr + elf.e_entry\n- else:\n- eentry = elf.e_entry\n+ eentry = baseoff + elf.e_entry\nif vw.isValidPointer(eentry):\nvw.addExport(eentry, EXP_FUNCTION, '__entry', fname)\n@@ -671,7 +661,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nreturn fname\n-def applyRelocs(elf, vw, addbase=False, baseaddr=0):\n+def applyRelocs(elf, vw, addbase=False, baseoff=0):\n'''\nprocess relocations / strings (relocs use Dynamic Symbols)\n'''\n@@ -682,8 +672,7 @@ def applyRelocs(elf, vw, addbase=False, baseaddr=0):\nfor r in relocs:\nrtype = Elf.getRelocType(r.r_info)\nrlva = r.r_offset\n- if addbase:\n- rlva += baseaddr\n+ rlva += baseoff\ntry:\n# If it has a name, it's an externally resolved \"import\" entry,\n# otherwise, just a regular reloc\n@@ -811,8 +800,7 @@ def applyRelocs(elf, vw, addbase=False, baseaddr=0):\nvw.makeName(rlva, pname)\n# name the target as well\n- if addbase:\n- ptr += baseaddr\n+ ptr += baseoff\n# normalize thumb addresses\nptr &= -2\nvw.makeName(ptr, name)\n@@ -839,8 +827,7 @@ def applyRelocs(elf, vw, addbase=False, baseaddr=0):\nif vw.vaByName(pname) is None:\nvw.makeName(rlva, pname)\n- if addbase:\n- ptr += baseaddr\n+ ptr += baseoff\nvw.makeImport(ptr, \"*\", dmglname)\nvw.setComment(ptr, name)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/teststorage.py",
"new_path": "vivisect/tests/teststorage.py",
"diff": "@@ -65,8 +65,8 @@ class StorageTests(unittest.TestCase):\nold = list(vw.exportWorkspace())\nnew = list(ovw.exportWorkspace())\n- self.assertEqual(len(old), 35)\n- self.assertEqual(len(new), 36) # the last event is a setMeta made by loadWorkspace\n+ self.assertEqual(len(old), 36)\n+ self.assertEqual(len(new), 37) # the last event is a setMeta made by loadWorkspace\nself.assertEqual(new[-1], (VWE_SETMETA, ('StorageName', self.tmpf.name)))\nfor idx in range(len(old)):\nself.assertEqual(old[idx], new[idx])\n@@ -96,14 +96,14 @@ class StorageTests(unittest.TestCase):\nmvw._event_list = []\nmvw.loadWorkspace(mpfile.name)\nmevt = list(mvw.exportWorkspace())\n- self.assertEqual(len(mevt), 36)\n+ self.assertEqual(len(mevt), 37)\nbvw = vivisect.VivWorkspace()\nbvw.setMeta('StorageModule', 'vivisect.storage.basicfile')\nbvw._event_list = []\nbvw.loadWorkspace(basicfile.name)\nbevt = list(bvw.exportWorkspace())\n- self.assertEqual(len(bevt), 36)\n+ self.assertEqual(len(bevt), 37)\n# the last three events are specific to the different storage modules\nfor idx in range(len(mevt) - 3):\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -191,7 +191,7 @@ class VivisectTest(v_t_utils.VivTest):\nself.chgrp_vw.do_filemeta('chgrp')\noutput = self.chgrp_vw.canvas.strval\nself.assertIn(\"'DT_INIT': 134516068\", output)\n- self.assertIn(\"'addbase': False,\\n 'canaries': False,\\n 'imagebase': 134512640\", output)\n+ self.assertIn(\"'canaries': False,\\n 'imagebase': 134512640\", output)\nself.chgrp_vw.canvas.clearCanvas()\ndef test_cli_fscope(self):\n@@ -717,7 +717,7 @@ class VivisectTest(v_t_utils.VivTest):\n'''\nvw = self.firefox_vw\nself.assertIsNotNone(vw.parsedbin)\n- self.assertEqual(set(['Emulation Anomalies', 'EntryPoints', 'SwitchCases', 'EmucodeFunctions', 'PointersFromFile', 'FuncWrappers', 'CodeFragments', 'DynamicBranches', 'Bookmarks', 'NoReturnCalls', 'DelayImports', 'Library Loads', 'pe:ordinals']), set(vw.getVaSetNames()))\n+ self.assertEqual(set(['Emulation Anomalies', 'EntryPoints', 'SwitchCases', 'EmucodeFunctions', 'PointersFromFile', 'FuncWrappers', 'CodeFragments', 'DynamicBranches', 'Bookmarks', 'NoReturnCalls', 'DelayImports', 'Library Loads', 'pe:ordinals', 'ResolvedImports']), set(vw.getVaSetNames()))\nself.assertEqual((0x14001fa5a, 6, 10, None), vw.getPrevLocation(0x14001fa60))\nself.assertEqual((0x14001fa5a, 6, 10, None), vw.getPrevLocation(0x14001fa60, adjacent=True))\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/vivbin.py",
"new_path": "vivisect/vivbin.py",
"diff": "@@ -44,6 +44,8 @@ def main():\nhelp='Path to a directory to use for config data')\nparser.add_argument('-a', '--autosave', dest='autosave', default=False, action='store_true',\nhelp='Autosave configuration data')\n+ parser.add_argument('-o', '--outfile', default=None,\n+ help='Name of VivWorkspace file to create (useful for loading multiple binaries into one workspace)')\nparser.add_argument('file', nargs='*')\nargs = parser.parse_args()\n@@ -54,6 +56,9 @@ def main():\nlevel = e_common.LOG_LEVELS[vw.verbose]\ne_common.initLogging(logger, level=level)\n+ if args.outfile:\n+ vw.setMeta('StorageName', args.outfile)\n+\n# do things\nif args.option is not None:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
feature: -o to select the output filename
bugfix: libc_start_main() errored out with a string mismatch
bugfix: move "connectExportsToImports()" to it's own "linker" analysis module
bugfix: ELF mobile loading ability!!!!!!!!!!!! <AGCK!>
feature: new VaSet "ResolvedImports" for imports that have been resolved and turned into Pointers.
|
718,770 |
23.11.2021 23:01:03
| 18,000 |
038e15410bd6c0b72d4a245043320617c00a7324
|
improvement: Dynamics add-base-offset before storage, and cleanup
unittest: adding multi-file and non-caching options in helpers. tests to come.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/elfplt.py",
"new_path": "vivisect/analysis/elf/elfplt.py",
"diff": "@@ -102,12 +102,6 @@ def getGOTByFilename(vw, filename):\nif fdyns is not None:\nFGOT = fdyns.get('DT_PLTGOT')\nif FGOT is not None:\n- # be sure to add the imgbase to FGOT if required\n- if vw.getFileMeta(filename, 'addbase'):\n- imgbase = vw.getFileMeta(filename, 'baseoff')\n- FGOT += imgbase\n- logger.debug('Adding Imagebase(%r) for Dynamics GOT: 0x%x (0x%x)', filename, imgbase, FGOT)\n-\nif FGOT != gotva and None not in (FGOT, gotva):\nlogger.warning(\"Dynamics and Sections have different GOT entries: S:0x%x D:0x%x. using Dynamics\", gotva, FGOT)\n@@ -135,12 +129,6 @@ def getGOTs(vw):\nif fdyns is not None:\nFGOT = fdyns.get('DT_PLTGOT')\nif FGOT is not None:\n- # be sure to add the imgbase to FGOT if required\n- if vw.getFileMeta(filename, 'addbase'):\n- imgbase = vw.getFileMeta(filename, 'baseoff')\n- FGOT += imgbase\n- logger.debug('Adding Imagebase(%r) for Dynamics GOT: 0x%x (0x%x)', filename, imgbase, FGOT)\n-\nflist = out[filename]\nskip = False\n@@ -177,11 +165,6 @@ def getPLTs(vw):\nif None in (FPLT, FPLTSZ):\ncontinue\n- if vw.getFileMeta(fname, 'addbase'):\n- imgbase = vw.getFileMeta(fname, 'baseoff')\n- FPLT += imgbase\n- logger.debug('Adding Imagebase(%r) for Dynamics PLT: 0x%x (0x%x)', fname, imgbase, FPLT)\n-\nnewish = True\nfor pltva, pltsize in plts:\nif FPLT == pltva:\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -234,6 +234,20 @@ def getAddBaseAddr(elf, baseaddr=None):\nlogger.debug(\"baseaddr: %r\\tbaseoff: 0x%x\", baseaddr, baseoff)\nreturn addbase, baseoff, baseaddr\n+# which Dynamic Types require rebasing\n+dt_rebase = (\n+ Elf.DT_INIT,\n+ Elf.DT_FINI,\n+ Elf.DT_INIT_ARRAY,\n+ Elf.DT_FINI_ARRAY,\n+ Elf.DT_GNU_HASH,\n+ Elf.DT_STRTAB,\n+ Elf.DT_SYMTAB,\n+ Elf.DT_PLTGOT,\n+ Elf.DT_JMPREL,\n+ Elf.DT_REL,\n+ Elf.DT_VERNEED,\n+)\ndef loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# analysis of discovered functions and data locations should be stored until the end of loading\nlogger.info(\"loadElfIntoWorkspace(filename=%r, baseaddr: %r\", filename, baseaddr)\n@@ -458,7 +472,11 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nelse:\nlogger.debug(\"DYNAMIC:\\t%r\", d)\n- elfmeta[Elf.dt_names.get(d.d_tag)] = d.d_value\n+ dval = d.d_value\n+ if d.d_tag in dt_rebase and addbase:\n+ dval += baseoff\n+\n+ elfmeta[Elf.dt_names.get(d.d_tag)] = dval\n# TODO: create a VaSet instead? setMeta allows more free-form info,\n# but isn't currently accessible from the gui\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/helpers.py",
"new_path": "vivisect/tests/helpers.py",
"diff": "@@ -32,13 +32,32 @@ def getTestPath(*paths):\n@functools.lru_cache()\n-def getTestWorkspace(*paths):\n+def getTestWorkspace(*paths, vw=None):\ntestdir = os.getenv('VIVTESTFILES')\nif not testdir:\nraise unittest.SkipTest('VIVTESTFILES env var not found!')\n+\n+ testdir = os.path.abspath(testdir)\n+ fpath = os.path.join(testdir, *paths)\n+\n+ if not vw:\n+ vw = v_cli.VivCli()\n+\n+ vw.loadFromFile(fpath)\n+ vw.analyze()\n+ return vw\n+\n+def getTestWorkspace_nocache(*paths, vw=None):\n+ testdir = os.getenv('VIVTESTFILES')\n+ if not testdir:\n+ raise unittest.SkipTest('VIVTESTFILES env var not found!')\n+\ntestdir = os.path.abspath(testdir)\nfpath = os.path.join(testdir, *paths)\n+\n+ if not vw:\nvw = v_cli.VivCli()\n+\nvw.loadFromFile(fpath)\nvw.analyze()\nreturn vw\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
improvement: Dynamics add-base-offset before storage, and cleanup
unittest: adding multi-file and non-caching options in helpers. tests to come.
|
718,770 |
24.11.2021 20:54:53
| 18,000 |
ae1922739af7e31353d38034a76dc71f4b3be41b
|
completed unittests for loading multiple binaries.
|
[
{
"change_type": "ADD",
"old_path": null,
"new_path": "vivisect/tests/testvivisect_multi.py",
"diff": "+import logging\n+import unittest\n+import vivisect.tests.helpers as helpers\n+import vivisect.tests.utils as v_t_utils\n+\n+\n+logger = logging.getLogger(__name__)\n+\n+\n+class VivisectMultiFileTest(v_t_utils.VivTest):\n+ def test_multiple_i386(self):\n+ vw = helpers.getTestWorkspace_nocache('linux', 'i386', 'chgrp.llvm')\n+ vw = helpers.getTestWorkspace_nocache('linux', 'i386', 'vdir.llvm', vw=vw)\n+ vw = helpers.getTestWorkspace_nocache('linux', 'i386', 'ld-2.31.so', vw=vw)\n+\n+ # test memory maps are not overlapping\n+ for mmap1 in vw.getMemoryMaps():\n+ for mmap2 in vw.getMemoryMaps():\n+ if mmap1 == mmap2:\n+ continue\n+\n+ mmva1, mmsz1, mmp1, mmnm1 = mmap1\n+ mmva2, mmsz2, mmp2, mmnm2 = mmap2\n+ self.assertFalse(mmva1 <= mmva2 < (mmva1 + mmsz1))\n+ self.assertFalse(mmva2 <= mmva1 < (mmva2 + mmsz2))\n+\n+ # test __entry values\n+ vdir = vw.parseExpression('vdir')\n+ chgrp = vw.parseExpression('chgrp')\n+ ld_2_31 = vw.parseExpression('ld_2_31')\n+\n+ vdir_entry = vw.parseExpression('vdir.__entry')\n+ chgrp_entry = vw.parseExpression('chgrp.__entry')\n+ ld_2_31_entry = vw.parseExpression('ld_2_31.__entry')\n+\n+ self.assertEqual(vdir_entry-vdir, 7024)\n+ self.assertEqual(chgrp_entry-chgrp, 4576)\n+ self.assertEqual(ld_2_31_entry-ld_2_31, 4384)\n+\n+ # test main functions\n+ vdir_main = vw.parseExpression('vdir.main')\n+ chgrp_main = vw.parseExpression('chgrp.main')\n+\n+ self.assertEqual(vdir_main-vdir, 7312)\n+ self.assertEqual(chgrp_main-chgrp, 5712)\n+\n+ # test PLT functions\n+ vdir_plt_names = [(va,name) for va,name in vw.getNames() if 'vdir.plt_' in name]\n+ chgrp_plt_names = [(va,name) for va,name in vw.getNames() if 'chgrp.plt_' in name]\n+ ld_2_31_plt_names = [(va,name) for va,name in vw.getNames() if 'ld_2_31.plt_' in name]\n+ self.assertEqual(len(vdir_plt_names), 0) # can't expect PLT analysis to work for a binary not intended to be relocated.\n+ self.assertEqual(len(chgrp_plt_names), 70) # this is more than we get when loading chgrp on it's own!?\n+ self.assertEqual(len(ld_2_31_plt_names), 8)\n+\n+ ## now just poke at one or two in each proggy\n+ ### chgrp __gmon_start__\n+ got_gmon_start = vw.parseExpression('chgrp+0xbffc')\n+ plt_gmon_start = vw.parseExpression('chgrp+0x11d0')\n+ self.assertIn('*.__gmon_start___', vw.getName(got_gmon_start))\n+ self.assertIn('plt___gmon_start__', vw.getName(plt_gmon_start))\n+ self.assertIn((plt_gmon_start, got_gmon_start, 2, 0), vw.getXrefsFrom(plt_gmon_start))\n+\n+ ### chgrp __assert_fail\n+ got_gmon_start = vw.parseExpression('chgrp+0xbffc')\n+ plt_gmon_start = vw.parseExpression('chgrp+0x11d0')\n+ self.assertIn('*.__gmon_start___', vw.getName(got_gmon_start))\n+ self.assertIn('plt___gmon_start__', vw.getName(plt_gmon_start))\n+ self.assertIn((plt_gmon_start, got_gmon_start, 2, 0), vw.getXrefsFrom(plt_gmon_start))\n+\n+ # test exports\n+ for exp in vw.getExports():\n+ eva, etype, esym, efile = exp\n+ mmap = vw.getMemoryMap(eva)\n+ self.assertEqual(mmap[3], efile, msg=\"Failed: %r for %r\" % (exp, mmap))\n+\n+ # test GOT locations\n+ vdir_main = vw.parseExpression('vdir.main')\n+ chgrp_main = vw.parseExpression('chgrp.main')\n+\n+ self.assertEqual(vdir_main-vdir, 7312)\n+ self.assertEqual(chgrp_main-chgrp, 5712)\n+\n+ # test Relocs\n+ self.assertIn(('ld_2_31', 179948, 2, 135749), vw.getRelocations())\n+ rel0 = vw.parseExpression('ld_2_31 + 179948')\n+ ptr0 = vw.parseExpression('ld_2_31 + 135749')\n+ self.assertIn((rel0, ptr0, 3, 0), vw.getXrefsFrom(rel0))\n+\n+\n+ # test Sections\n+ for seg in vw.getSegments():\n+ if seg[2] == '.text':\n+ if seg[3] == 'vdir':\n+ self.assertEqual(7024, seg[0] - vw.parseExpression('vdir'))\n+ if seg[3] == 'chgrp':\n+ self.assertEqual(4576, seg[0] - vw.parseExpression('chgrp'))\n+ if seg[3] == 'ld_2_31':\n+ self.assertEqual(4352, seg[0] - vw.parseExpression('ld_2_31'))\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
completed unittests for loading multiple binaries.
|
718,770 |
01.12.2021 12:41:11
| 18,000 |
8a305492724aef55bcaaef4e4a201432a2c2f1d5
|
a couple related bugfixes found floating around.
|
[
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -574,7 +574,7 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\nif dohex:\nmemstr = binascii.unhexlify(memstr)\nif douni:\n- memstr = (\"\\x00\".join(memstr)) + \"\\x00\"\n+ memstr = (b\"\\x00\".join(memstr)) + b\"\\x00\"\naddr = self.parseExpression(exprstr)\nself.memobj.writeMemory(addr, memstr)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -566,11 +566,12 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nElf.st_info_bind.get(s.st_other, s.st_other),\ns.name)\n- if s.getInfoType() == Elf.STT_FILE:\n+ symtype = s.getInfoType()\n+ if symtype == Elf.STT_FILE:\nvw.setVaSetRow('FileSymbols', (dmglname, sva))\ncontinue\n- elif s.getInfoType() == Elf.STT_NOTYPE:\n+ elif symtype == Elf.STT_NOTYPE:\n# mapping symbol\nif arch in ('arm', 'thumb', 'thumb16'):\nsymname = s.getName()\n@@ -589,7 +590,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# Data Items (eg. literal pool)\nlogger.info('mapping (NOTYPE) data symbol: 0x%x: %r', sva, dmglname)\ndata_ptrs.append(sva)\n- elif s.getInfoType() == Elf.STT_OBJECT:\n+ elif symtype == Elf.STT_OBJECT:\nsymname = s.getName()\nsva += baseoff\nif symname:\n@@ -622,6 +623,15 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nelse:\nvw.makeNumber(sva, size=s.st_size)\n+ elif symtype == Elf.STT_FUNC:\n+ sva += baseoff\n+\n+ symname = s.getName()\n+ if symname:\n+ vw.makeName(sva, symname, filelocal=True, makeuniq=True)\n+\n+ new_functions.append((\"Symbol: FUNC\", sva))\n+\n# if the symbol has a value of 0, it is likely a relocation point which gets updated\nsname = demangle(s.name)\nif sva == 0:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
a couple related bugfixes found floating around.
|
718,770 |
01.12.2021 17:02:39
| 18,000 |
011897323f7e66280cb7aeeb1d1b66f053173c86
|
woops, make linker work again.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/generic/linker.py",
"new_path": "vivisect/analysis/generic/linker.py",
"diff": "@@ -3,5 +3,4 @@ Connect any Exports we have to Imports we may also have\n(most useful for multiple file workspaces)\n'''\ndef analyze(vw):\n- #vw.connectImportsWithExports()\n- pass\n+ vw.connectImportsWithExports()\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
woops, make linker work again.
|
718,770 |
01.12.2021 19:04:08
| 18,000 |
2a3dcbdc428e90e47a61c81e4cc78f04f63d7aa7
|
kill it. too many issues
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -623,15 +623,6 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nelse:\nvw.makeNumber(sva, size=s.st_size)\n- elif symtype == Elf.STT_FUNC:\n- sva += baseoff\n-\n- symname = s.getName()\n- if symname and vw.isValidPointer(sva):\n- vw.makeName(sva, symname, filelocal=True, makeuniq=True)\n-\n- new_functions.append((\"Symbol: FUNC\", sva))\n-\n# if the symbol has a value of 0, it is likely a relocation point which gets updated\nsname = demangle(s.name)\nif sva == 0:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
kill it. too many issues
|
718,770 |
01.12.2021 23:34:38
| 18,000 |
9a5693908e1c238b9f2ec9b8845c4bc44d8d08b5
|
rearrange ELF analysis to catch function thunks (elfplt/entrypoints).
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2835,6 +2835,10 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nif impfname not in (efname, '*'):\ncontinue\n+ if self.isFunctionThunk(eva):\n+ logger.info(\"Skipping Exported Thunk\")\n+ continue\n+\n# file and symbol name match. apply the magic.\n# do we ever *not* write in the full address at the import site?\nlogger.debug(\"connecting Import 0x%x -> Export 0x%x (%r)\", iva, eva, isym)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/__init__.py",
"new_path": "vivisect/analysis/__init__.py",
"diff": "@@ -91,9 +91,9 @@ def addAnalysisModules(vw):\n# elfplt wants to be run before generic.entrypoints.\nvw.addAnalysisModule(\"vivisect.analysis.elf.elfplt\")\n- vw.addAnalysisModule(\"vivisect.analysis.generic.linker\")\nvw.addAnalysisModule(\"vivisect.analysis.generic.entrypoints\")\nvw.addAnalysisModule(\"vivisect.analysis.elf\")\n+ vw.addAnalysisModule(\"vivisect.analysis.generic.linker\")\nif arch in ('i386', 'amd64', 'arm'):\nvw.addImpApi('posix', arch)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
rearrange ELF analysis to catch function thunks (elfplt/entrypoints).
|
718,770 |
02.12.2021 00:06:25
| 18,000 |
f488325cf0f6d0915c17e7ce6808a5db01fabc61
|
even later for ELF. make sure all of PLT is analyzed before linking.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/__init__.py",
"new_path": "vivisect/analysis/__init__.py",
"diff": "@@ -93,7 +93,6 @@ def addAnalysisModules(vw):\nvw.addAnalysisModule(\"vivisect.analysis.elf.elfplt\")\nvw.addAnalysisModule(\"vivisect.analysis.generic.entrypoints\")\nvw.addAnalysisModule(\"vivisect.analysis.elf\")\n- vw.addAnalysisModule(\"vivisect.analysis.generic.linker\")\nif arch in ('i386', 'amd64', 'arm'):\nvw.addImpApi('posix', arch)\n@@ -140,6 +139,7 @@ def addAnalysisModules(vw):\nvw.addFuncAnalysisModule(\"vivisect.analysis.elf.elfplt\")\n# late-analysis ELF PLT tidying up, allowing unused PLT entries to be made into functions\nvw.addAnalysisModule(\"vivisect.analysis.elf.elfplt_late\")\n+ vw.addAnalysisModule(\"vivisect.analysis.generic.linker\")\nvw.addAnalysisModule(\"vivisect.analysis.generic.thunks\")\nvw.addAnalysisModule(\"vivisect.analysis.generic.pointers\")\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
even later for ELF. make sure all of PLT is analyzed before linking.
|
718,770 |
27.12.2021 17:05:57
| 18,000 |
b28b46c8fb302ba9b8e813179178e66e182c096d
|
updates to support big-endian MachO
|
[
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/macho/__init__.py",
"new_path": "vstruct/defs/macho/__init__.py",
"diff": "@@ -4,8 +4,6 @@ Structure definitions for the OSX MachO binary format.\nimport struct\nimport logging\nimport binascii\n-import vstruct\n-import envi.bits as e_bits\nimport vstruct\nimport envi.bits as e_bits\n"
},
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/macho/loader.py",
"new_path": "vstruct/defs/macho/loader.py",
"diff": "import vstruct\n-import envi.common as e_common\nfrom vstruct.primitives import *\nfrom vstruct.defs.macho.const import *\n@@ -502,9 +501,7 @@ class data_in_code_entry(vstruct.VStruct):\ncommand_classes = {\nLC_SEGMENT: segment_command,\n- LC_SEGMENT_64: segment_command_64,\nLC_SYMTAB: symtab_command,\n- LC_DYSYMTAB: dysymtab_command,\nLC_LOAD_DYLIB: dylib_command,\nLC_SEGMENT: segment_command,\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
updates to support big-endian MachO
|
718,770 |
27.12.2021 17:30:06
| 18,000 |
8f05afcd8cd9e97b2c4990db0413552089ae2a70
|
make LC_UNIXTHREAD/thread_command work with .flavor (and getEntryPoints())
updates to work with bigendian (and wonky git-shit)
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/macho.py",
"new_path": "vivisect/parsers/macho.py",
"diff": "@@ -130,8 +130,8 @@ def _loadMacho(vw, filebytes, filename=None, baseaddr=None):\nfor libname in macho.getLibDeps():\nlogger.debug(\"deplib: %r\", libname)\n# FIXME hack....\n- libname = libname.split(b'/')[-1]\n- libname = libname.split(b'.')[0].lower()\n+ libname = libname.split('/')[-1]\n+ libname = libname.split('.')[0].lower()\nvw.addLibraryDependancy(libname)\nlogger.debug(\"memoryMaps: \\n%r\", '\\n'.join([\"0x%x, 0x%x, 0x%x, %r\" % (w,x,y,z) for w,x,y,z in vw.getMemoryMaps()]))\n"
},
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/macho/loader.py",
"new_path": "vstruct/defs/macho/loader.py",
"diff": "@@ -234,12 +234,32 @@ class dylinker_command(load_command):\nload_command.__init__(self, bigend=bigend)\nself.name = lc_str() # dynamic linker's path name\n-\n+# Thread commands contain machine-specific data structures suitable for\n+# use in the thread state primitives. The machine specific data structures\n+# follow the struct thread_command as follows.\n+# Each flavor of machine specific data structure is preceded by an unsigned\n+# long constant for the flavor of that data structure, an uint32_t\n+# that is the count of longs of the size of the state data structure and then\n+# the state data structure follows. This triple may be repeated for many\n+# flavors. The constants for the flavors, counts and state data structure\n+# definitions are expected to be in the header file <machine/thread_status.h>.\n+# These machine specific data structures sizes must be multiples of\n+# 4 bytes The cmdsize reflects the total size of the thread_command\n+# and all of the sizes of the constants for the flavors, counts and state\n+# data structures.\n+#\n+# For executable objects that are unix processes there will be one\n+# thread_command (cmd == LC_UNIXTHREAD) created for it by the link-editor.\n+# This is the same as a LC_THREAD, except that a stack is automatically\n+# created (based on the shell's limit for the stack size). Command arguments\n+# and environment variables are copied onto that stack.\nclass thread_command(load_command):\ndef __init__(self, bigend=False):\n# LC_THREAD or LC_UNIXTHREAD\nload_command.__init__(self, bigend=bigend)\n+ self.flavor = v_uint32(bigend=bigend) # flavor of thread state\n+ self.count = v_uint32(bigend=bigend) # count of longs in thread state\nclass routines_command(load_command):\n@@ -467,6 +487,10 @@ class version_min_command(load_command):\nself.sdk = v_uint32(bigend=bigend)\n+# The entry_point_command is a replacement for thread_command.\n+# It is used for main executables to specify the location (file offset)\n+# of main(). If -stack_size was used at link time, the stacksize\n+# field will contain the stack size need for the main thread.\nclass entry_point_command(load_command):\ndef __init__(self, bigend=False):\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
make LC_UNIXTHREAD/thread_command work with .flavor (and getEntryPoints())
updates to work with bigendian (and wonky git-shit)
|
718,770 |
28.12.2021 09:05:50
| 18,000 |
c17e328a69baa17b86226d33f6714eebd0d96df2
|
bugfix: vsParse doesn't have bigend defined
|
[
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/macho/loader.py",
"new_path": "vstruct/defs/macho/loader.py",
"diff": "@@ -74,7 +74,7 @@ class segment_command(load_command):\n# segment, which follows directly after the segment info structure\nstep = len(self)\nfor i in range(self.nsects):\n- sect = section(bigend=bigend)\n+ sect = section(bigend=self.getEndian())\nstep += sect.vsParse(bytes[offset+step:])\nself.sections.vsAddElement(sect)\n@@ -106,7 +106,7 @@ class segment_command_64(load_command):\n# segment, which follows directly after the segment info structure\nstep = len(self)\nfor i in range(self.nsects):\n- sect = section_64(bigend=bigend)\n+ sect = section_64(self.getEndian())\nstep += sect.vsParse(bytes[offset+step:])\nself.sections.vsAddElement(sect)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
bugfix: vsParse doesn't have bigend defined
|
718,770 |
05.01.2022 22:27:22
| 18,000 |
cbd95278008751b9223c8911e08d2b5251514a3f
|
minor changes to PE parser to not overwrite existing VaSets
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/pe.py",
"new_path": "vivisect/parsers/pe.py",
"diff": "@@ -93,6 +93,7 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\nvw.setMeta('Architecture', arch)\nvw.setMeta('Format', 'pe')\nvw.parsedbin = pe\n+ ### TODO: this doesn't work from Memory (read() needs to know how big the file is)\nbyts = pe.getFileBytes()\nvw.setMeta('FileBytes', v_parsers.compressBytes(byts))\n@@ -164,8 +165,12 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\nvw.setFileMeta(fname, 'Version', vsver)\n# Setup some va sets used by windows analysis modules\n+ # these will already exist for Multi-file workspaces\n+ if not 'Library Loads' in vw.getVaSetNames():\nvw.addVaSet(\"Library Loads\", ((\"Address\", VASET_ADDRESS), (\"Library\", VASET_STRING)))\n+ if not 'pe:ordinals' in vw.getVaSetNames():\nvw.addVaSet('pe:ordinals', (('Address', VASET_ADDRESS), ('Ordinal', VASET_INTEGER)))\n+ if not 'DelayImports' in vw.getVaSetNames():\nvw.addVaSet('DelayImports', (('Address', VASET_ADDRESS), ('DelayImport', VASET_STRING)))\n# SizeOfHeaders spoofable...\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
minor changes to PE parser to not overwrite existing VaSets
|
718,765 |
12.01.2022 14:49:10
| 18,000 |
e4bc9b93cd00853aa9d6d981b185e086c5fc5e0f
|
more mach-o definitions and support
Add more definitions for FAT64 and CPU arch names/constants.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/macho.py",
"new_path": "vivisect/parsers/macho.py",
"diff": "import envi.common as e_common\nimport vivisect.parsers as viv_parsers\n+\nimport vstruct.defs.macho as vs_macho\n+import vstruct.defs.macho.fat as vsm_fat\n+import vstruct.defs.macho.const as vsm_const\ndef parseFile(vw, filename, baseaddr=None):\nwith open(filename, 'rb') as f:\n@@ -49,10 +52,15 @@ def _loadMacho(vw, filebytes, filename=None, baseaddr=None):\narchlist = []\noffset = 0\n- fat = vs_macho.fat_header()\n+ fat = vsm_fat.fat_header()\noffset = fat.vsParse(filebytes, offset=offset)\n+ if fat.magic in vsm_const.FAT_64:\n+ archcon = vsm_fat.fat_arch_64\n+ else:\n+ archcon = vsm_fat.fat_arch\n+\nfor i in range(fat.nfat_arch):\n- ar = vs_macho.fat_arch()\n+ ar = archcon()\noffset = ar.vsParse(filebytes, offset=offset)\narchname = vs_macho.mach_cpu_names.get(ar.cputype)\nif archname == fatarch:\n"
},
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/macho/const.py",
"new_path": "vstruct/defs/macho/const.py",
"diff": "@@ -130,21 +130,27 @@ S_ATTR_LOC_RELOC = 0x00000100 # section has local relocation entries\nINDIRECT_SYMBOL_LOCAL = 0x80000000 # section has local relocation entries\nINDIRECT_SYMBOL_ABS = 0x40000000 # section has local relocation entries\n+CPU_ARCH_MASK = 0xff000000\n+CPU_ARCH_ABI64 = 0x01000000\n+CPU_ARCH_ABI64_32 = 0x02000000\n+\nCPU_TYPE_ANY = -1\nCPU_TYPE_VAX = 1\nCPU_TYPE_MC680 = 6\nCPU_TYPE_X86 = 7\n-CPU_TYPE_X86_64 = 0x01000007\n+CPU_TYPE_X86_64 = CPU_TYPE_X86 | CPU_ARCH_ABI64\nCPU_TYPE_MIPS = 8\nCPU_TYPE_MC98000 = 10\nCPU_TYPE_HPPA = 11\nCPU_TYPE_ARM = 12\n+CPU_TYPE_ARM64 = CPU_TYPE_ARM | CPU_ARCH_ABI64\n+CPU_TYPE_ARM64_32 = CPU_TYPE_ARM | CPU_ARCH_ABI64_32\nCPU_TYPE_MC88000 = 13\nCPU_TYPE_SPARC = 14\nCPU_TYPE_I860 = 15\nCPU_TYPE_ALPHA = 16\nCPU_TYPE_POWERPC = 18\n-#CPU_TYPE_POWERPC64 (CPU_TYPE_POWERPC | CPU_ARCH_ABI64)\n+CPU_TYPE_POWERPC64 = CPU_TYPE_POWERPC | CPU_ARCH_ABI64\nmach_cpu_names = {\nCPU_TYPE_VAX : 'vax',\n@@ -155,11 +161,14 @@ mach_cpu_names = {\nCPU_TYPE_MC98000 : 'mc98000',\nCPU_TYPE_HPPA : 'hppa',\nCPU_TYPE_ARM : 'arm',\n+ CPU_TYPE_ARM64 : 'arm64',\n+ CPU_TYPE_ARM64_32 : 'arm64_32',\nCPU_TYPE_MC88000 : 'mc88000',\nCPU_TYPE_SPARC : 'sparc',\nCPU_TYPE_I860 : 'i860',\nCPU_TYPE_ALPHA : 'alpha',\nCPU_TYPE_POWERPC : 'powerpc',\n+ CPU_TYPE_POWERPC64 : 'powerpc64',\n}\n# Symbol types\n"
},
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/macho/fat.py",
"new_path": "vstruct/defs/macho/fat.py",
"diff": "@@ -18,3 +18,12 @@ class fat_arch(vstruct.VStruct):\nself.size = v_uint32(bigend=True) # size of this object file */\nself.align = v_uint32(bigend=True) # alignment as a power of 2 */\n+class fat_arch_64(vstruct.VStruct):\n+ def __init__(self):\n+ vstruct.VStruct.__init__(self)\n+ self.cputype = v_uint32(bigend=True) # cpu specifier (int) */\n+ self.cpusubtype = v_uint32(bigend=True) # machine specifier (int) */\n+ self.offset = v_uint64(bigend=True) # file offset to this object file */\n+ self.size = v_uint64(bigend=True) # size of this object file */\n+ self.align = v_uint32(bigend=True) # alignment as a power of 2 */\n+ self.reserved = v_uint32(bigend=True)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
more mach-o definitions and support (#495)
Add more definitions for FAT64 and CPU arch names/constants.
|
718,765 |
13.01.2022 10:52:06
| 18,000 |
4869e826c27f7f3cec5dd64169766f5edb6246bd
|
fix pe32 vs pe32+ check
Use the OptionalHeader magic instead of the FileHeader.Machine value for determining PE32 vs PE32+
|
[
{
"change_type": "MODIFY",
"old_path": "PE/__init__.py",
"new_path": "PE/__init__.py",
"diff": "@@ -10,6 +10,9 @@ import vivisect.exc as v_exc\nfrom . import ordlookup\n+PE32_MAGIC = 0x10b\n+PE32PLUS_MAGIC = 0x20b\n+\nIMAGE_FILE_RELOCS_STRIPPED = 0x0001\nIMAGE_FILE_EXECUTABLE_IMAGE = 0x0002\nIMAGE_FILE_LINE_NUMS_STRIPPED = 0x0004\n@@ -406,14 +409,16 @@ class PE(object):\ndosbytes = self.readAtOffset(0, len(self.IMAGE_DOS_HEADER))\nself.IMAGE_DOS_HEADER.vsParse(dosbytes)\n- nt = self.readStructAtOffset(self.IMAGE_DOS_HEADER.e_lfanew, \"pe.IMAGE_NT_HEADERS\")\n-\n# Parse in a default 32 bit, and then check for 64...\n- if nt.FileHeader.Machine in [ IMAGE_FILE_MACHINE_AMD64, IMAGE_FILE_MACHINE_IA64 ]:\n+ nt = self.readStructAtOffset(self.IMAGE_DOS_HEADER.e_lfanew, \"pe.IMAGE_NT_HEADERS\")\n+ magic = struct.unpack(\"<H\", nt.OptionalHeader.Magic)[0]\n+ if magic == PE32PLUS_MAGIC:\nnt = self.readStructAtOffset(self.IMAGE_DOS_HEADER.e_lfanew, \"pe.IMAGE_NT_HEADERS64\")\nself.pe32p = True\nself.psize = 8\nself.high_bit_mask = 0x8000000000000000\n+ elif magic != PE32_MAGIC:\n+ logger.warning('nt.OptionalHeader magic got invalid value of %x', magic)\nself.IMAGE_NT_HEADERS = nt\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testpe.py",
"new_path": "vivisect/tests/testpe.py",
"diff": "@@ -7,12 +7,14 @@ import envi.memory as e_memory\nimport vivisect\nimport vivisect.const as viv_con\nimport vivisect.parsers.pe\n+\nimport vivisect.tests.helpers as helpers\n+import vivisect.tests.utils as v_t_utils\nlogger = logging.getLogger(__name__)\n-class PETests(unittest.TestCase):\n+class PETests(v_t_utils.VivTest):\n@classmethod\ndef setUpClass(cls):\n@@ -508,7 +510,7 @@ class PETests(unittest.TestCase):\n(4371200, 4, 9, 'gdi32.StartPage')])\nfailed = False\n- # self.assertEqual(imps, vw.getImports())\n+\nfor iloc in vw.getImports():\nif iloc not in imps:\nlogger.critical(f'{iloc} was not found in the predefined imports. A new one?')\n@@ -518,10 +520,36 @@ class PETests(unittest.TestCase):\nfor iloc in imps:\nif iloc not in oimp:\nlogger.critical(f'The imports are missing {iloc}.')\n+ failed = True\nif failed:\nself.fail('Please see test logs for import test failures')\n+ def test_pe_pe32p(self):\n+ '''\n+ test that we use the right check for constructing the optional header\n+ and that the data directories are right\n+ '''\n+ file_path = helpers.getTestPath('windows', 'amd64', 'a7712b7c45ae081a1576a387308077f808c666449d1ea9ba680ec410569d476f.file')\n+ pe = PE.peFromFileName(file_path)\n+ self.assertTrue(pe.pe32p)\n+ self.eq(pe.psize, 8)\n+\n+ self.eq(pe.IMAGE_NT_HEADERS.OptionalHeader.SizeOfStackReserve, 0x100000)\n+ self.eq(pe.IMAGE_NT_HEADERS.OptionalHeader.SizeOfStackCommit, 0x1000)\n+ self.eq(pe.IMAGE_NT_HEADERS.OptionalHeader.SizeOfHeapReserve, 0x100000)\n+ self.eq(pe.IMAGE_NT_HEADERS.OptionalHeader.SizeOfHeapCommit, 0x1000)\n+ self.eq(pe.IMAGE_NT_HEADERS.OptionalHeader.LoaderFlags, 0)\n+ self.eq(pe.IMAGE_NT_HEADERS.OptionalHeader.NumberOfRvaAndSizes, 0x10)\n+\n+ ddir = pe.getDataDirectory(PE.IMAGE_DIRECTORY_ENTRY_COM_DESCRIPTOR)\n+ self.eq(ddir.VirtualAddress, 0x2008)\n+ self.eq(ddir.Size, 0x48)\n+\n+ ddir = pe.getDataDirectory(PE.IMAGE_DIRECTORY_ENTRY_RESOURCE)\n+ self.eq(ddir.VirtualAddress, 0x9a2000)\n+ self.eq(ddir.Size, 0x3c78)\n+\ndef test_hiaddr_imports(self):\n# Test if imports located at a high relative address are discovered.\nfile_path = helpers.getTestPath('windows', 'i386', 'section_has_hi_virtualaddr.exe')\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
fix pe32 vs pe32+ check (#494)
Use the OptionalHeader magic instead of the FileHeader.Machine value for determining PE32 vs PE32+
|
718,765 |
13.01.2022 11:22:37
| 18,000 |
37b0b655d8dedfcf322e86b0f144b096e48d547e
|
V1.0.7 Release Prep
v1.0.7 release prep
|
[
{
"change_type": "MODIFY",
"old_path": ".bumpversion.cfg",
"new_path": ".bumpversion.cfg",
"diff": "[bumpversion]\n-current_version = 1.0.6\n+current_version = 1.0.7\n[bumpversion:file:setup.py]\nsearch = VERSION = '{current_version}'\n"
},
{
"change_type": "MODIFY",
"old_path": "CHANGELOG.rst",
"new_path": "CHANGELOG.rst",
"diff": "Vivisect Changelog\n******************\n+V1.0.7 - 2022-01-13\n+===================\n+\n+Features\n+--------\n+- More Mach-O structure definitions and parsing support.\n+ (`#495 <https://github.com/vivisect/vivisect/pull/495>`_)\n+\n+Fixes\n+-----\n+- Tweak how i386 analysis detections calling conventions.\n+ (`#493 <https://github.com/vivisect/vivisect/pull/493>`_)\n+- Use OptionalHeader.Magic for determining PE32/PE32+.\n+ (`#494 <https://github.com/vivisect/vivisect/pull/494>`_)\n+\nV1.0.6 - 2022-01-03\n===================\n"
},
{
"change_type": "MODIFY",
"old_path": "setup.py",
"new_path": "setup.py",
"diff": "from setuptools import find_packages, setup\nfrom os import path\n-VERSION = '1.0.6'\n+VERSION = '1.0.7'\ndirn = path.abspath(path.dirname(__file__))\nwith open(path.join(dirn, 'README.md'), 'r') as fd:\n"
},
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -2299,6 +2299,6 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these yourself\n-version = (1, 0, 6)\n+version = (1, 0, 7)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -3168,6 +3168,6 @@ def getVivPath(*pathents):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these directly\n-version = (1, 0, 6)\n+version = (1, 0, 7)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
V1.0.7 Release Prep (#496)
v1.0.7 release prep
|
718,770 |
05.01.2022 22:37:33
| 18,000 |
bd0a1266fc3f0e7f2c42e27a7650caacd1548a7a
|
improving the MemObjFile shim to make a memobj act like a file object.
|
[
{
"change_type": "MODIFY",
"old_path": "vstruct/__init__.py",
"new_path": "vstruct/__init__.py",
"diff": "@@ -20,7 +20,16 @@ class MemObjFile:\ndef seek(self, offset):\nself.offset = self.baseaddr + offset\n- def read(self, size):\n+ def flush(self):\n+ pass\n+\n+ def tell(self):\n+ return self.offset - self.baseaddr\n+\n+ def read(self, size=None):\n+ if size is None:\n+ # end of map for now, but perhaps this should be end of contiguous maps\n+ _, size, _, _ = self.memobj.getMemoryMap(self.offset)\nret = self.memobj.readMemory(self.offset, size)\nself.offset += size\nreturn ret\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
improving the MemObjFile shim to make a memobj act like a file object.
|
718,770 |
19.01.2022 23:26:59
| 18,000 |
9fb4b35a102b1ab42dc2ec52acc380c3bb6551fb
|
we recreate the workspace through the proxy, but never snap in the appropriate analysis modules. this likely dates back to the days when analysis modules were added to the eventlist.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -630,6 +630,7 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nself.server.vprint('%s connecting...' % uname)\nwsevents = self.server.exportWorkspace()\nself.importWorkspace(wsevents)\n+ self._snapInAnalysisModules()\nself.server.vprint('%s connection complete!' % uname)\nthr = threading.Thread(target=self._clientThread)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
we recreate the workspace through the proxy, but never snap in the appropriate analysis modules. this likely dates back to the days when analysis modules were added to the eventlist. (#498)
|
718,770 |
10.02.2022 21:08:36
| 18,000 |
73e2fb65f8a4e5531189d1f23974ba6caf0815cf
|
fix a bug in elfplt for an odd binary.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/elfplt.py",
"new_path": "vivisect/analysis/elf/elfplt.py",
"diff": "@@ -172,6 +172,9 @@ def getPLTs(vw):\nif fdyns is not None:\nFGOT = fdyns.get('DT_JMPREL')\nFGOTSZ = fdyns.get('DT_PLTRELSZ')\n+ if None in (FGOT, FGOTSZ):\n+ continue\n+\nif vw.getFileMeta(fname, 'addbase'):\nimgbase = vw.getFileMeta(fname, 'imagebase')\nlogger.debug('Adding Imagebase: 0x%x', imgbase)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
fix a bug in elfplt for an odd binary. (#503)
|
718,770 |
04.03.2022 23:23:17
| 18,000 |
931708510e5d725768a87ef89805e1be6c07075a
|
save bugfix
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/qt/main.py",
"new_path": "vivisect/qt/main.py",
"diff": "@@ -487,7 +487,7 @@ class VQVivMainWindow(viv_base.VivEventDist, vq_app.VQMainCmdWindow):\n# forces a local save\nfname = filename\nif not fname:\n- fname = vw.getMeta(\"StorageName\")\n+ fname = self.vw.getMeta(\"StorageName\")\nself.vw.vprint('Saving workspace... (%r)' % fname)\ntry:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
save bugfix (#507)
|
718,770 |
16.03.2022 16:29:38
| 14,400 |
05baf6b375e0d031a25b04a8827aa8304ab1bd76
|
make the loader use 1M alignment by default (and make it settable, instead of hard-coding it to ENVI's page-size)
|
[
{
"change_type": "MODIFY",
"old_path": "envi/const.py",
"new_path": "envi/const.py",
"diff": "@@ -41,4 +41,3 @@ for idx in range(8):\nPAGE_SIZE = 1 << 12\nPAGE_NMASK = PAGE_SIZE - 1\nPAGE_MASK = ~PAGE_NMASK\n-\n"
},
{
"change_type": "MODIFY",
"old_path": "envi/memory.py",
"new_path": "envi/memory.py",
"diff": "@@ -440,7 +440,7 @@ class MemoryObject(IMemory):\nself.addMemoryMap(baseva, perms, name, fill*size, align)\nreturn baseva\n- def findFreeMemoryBlock(self, size, suggestaddr=0x1000, MIN_MEM_ADDR = 0x1000):\n+ def findFreeMemoryBlock(self, size, suggestaddr=0x1000, MIN_MEM_ADDR = 0x1000, mapalign=0x10000):\n'''\nFind a block of memory in the address-space of the correct size which\ndoesn't overlap any existing maps. Attempts to offer the map starting\n@@ -456,6 +456,9 @@ class MemoryObject(IMemory):\ntmpva = suggestaddr\nmaxaddr = (1 << (8 * self.imem_psize)) - 1\n+ MAP_ALIGN_NMASK = mapalign - 1\n+ MAP_ALIGN_MASK = ~MAP_ALIGN_NMASK\n+\nwhile baseva is None:\n# if we roll into illegal memory, start over at page 2. skip 0.\nif tmpva > maxaddr:\n@@ -478,8 +481,8 @@ class MemoryObject(IMemory):\n# we ran into a memory map. adjust.\ngood = False\ntmpva = mmendva\n- tmpva += PAGE_NMASK\n- tmpva &= PAGE_MASK\n+ tmpva += MAP_ALIGN_NMASK\n+ tmpva &= MAP_ALIGN_MASK\nbreak\nif good:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
make the loader use 1M alignment by default (and make it settable, instead of hard-coding it to ENVI's page-size)
|
718,765 |
30.03.2022 15:05:47
| 14,400 |
a3333bd02e9383a3868bf614dc09782b33dc7de2
|
tweak how we handle section names
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testpe.py",
"new_path": "vivisect/tests/testpe.py",
"diff": "@@ -595,3 +595,21 @@ class PETests(v_t_utils.VivTest):\nself.assertEqual(loc[1], 40)\nself.assertEqual(loc[2], viv_con.LOC_STRUCT)\nself.assertEqual(loc[3], 'pe.IMAGE_SECTION_HEADER')\n+\n+ def test_bad_pe_sectname(self):\n+ vw = helpers.getTestWorkspace('windows', 'i386', 'HelloSection-err.exe')\n+ segs = vw.getSegments()\n+ pesections = [\n+ # about as blunt and direct of a name as we can get\n+ \"[invalid name] b'\\\\xfftext\\\\x00\\\\x00\\\\x00'\",\n+ '.rdata\\x00\\x00',\n+ '.data\\x00\\x00\\x00',\n+ '.rsrc\\x00\\x00\\x00',\n+ '.reloc\\x00\\x00',\n+ ]\n+ sections = [x.Name for x in vw.parsedbin.getSections()]\n+ valids = [x.isNameValid() for x in vw.parsedbin.getSections()]\n+ self.eq(valids, [False, True, True, True, True])\n+ self.len(sections, 5)\n+ for name in pesections:\n+ self.isin(name, sections)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/utils.py",
"new_path": "vivisect/tests/utils.py",
"diff": "@@ -33,6 +33,12 @@ class VivTest(unittest.TestCase):\n'''\nself.assertIsNotNone(x)\n+ def isin(self, x, y):\n+ '''\n+ Assert that x is a member of the container y\n+ '''\n+ self.assertIn(x, y)\n+\n@contextlib.contextmanager\ndef snap(self, vw):\n'''\n"
},
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/pe.py",
"new_path": "vstruct/defs/pe.py",
"diff": "+import binascii\nimport vstruct\nfrom vstruct.primitives import *\n@@ -264,11 +265,43 @@ class VS_FIXEDFILEINFO(vstruct.VStruct):\nself.FileDateMS = v_uint32()\nself.FileDateLS = v_uint32()\n+class v_name(v_bytes):\n+ '''\n+ Specifically for things that *should* be utf-8 strings, but aren't really enforced as such\n+ '''\n+ def __repr__(self):\n+ try:\n+ name = self._vs_value.decode('utf-8')\n+ return name\n+ except UnicodeDecodeError:\n+ return binascii.hexlify(self._vs_value).decode('utf-8')\n+\n+ __str__ = __repr__\n+\n+ def vsGetValue(self):\n+ '''\n+ Get the type specific value for this field.\n+ (Used by the structure dereference method to return\n+ a python native for the field by name)\n+ '''\n+ try:\n+ name = self._vs_value.decode('utf-8')\n+ return name\n+ except UnicodeDecodeError:\n+ return \"[invalid name] %r\" % self._vs_value\n+\n+ def isValid(self):\n+ try:\n+ self._vs_value.decode('utf-8')\n+ return True\n+ except UnicodeDecodeError:\n+ return False\n+\nclass IMAGE_SECTION_HEADER(vstruct.VStruct):\ndef __init__(self):\nvstruct.VStruct.__init__(self)\n- self.Name = v_str(8)\n+ self.Name = v_name(8)\nself.VirtualSize = v_uint32()\nself.VirtualAddress = v_uint32()\nself.SizeOfRawData = v_uint32()\n@@ -279,6 +312,10 @@ class IMAGE_SECTION_HEADER(vstruct.VStruct):\nself.NumberOfLineNumbers = v_uint16()\nself.Characteristics = v_uint32()\n+ def isNameValid(self):\n+ # we've gotta bypass\n+ name = self.vsGetField('Name')\n+ return name.isValid()\nclass IMAGE_RUNTIME_FUNCTION_ENTRY(vstruct.VStruct):\n\"\"\"\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
tweak how we handle section names (#514)
|
718,770 |
14.04.2022 13:51:09
| 14,400 |
9b651db3606956e104a32cfade96842670301a70
|
improved output from the do_names command.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2323,6 +2323,23 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nreturn foff\nreturn (foff - offset) + 2\n+ def getFileAndOffset(self, va):\n+ '''\n+ Helper function which identifies the file a given VA is a part of, then\n+ splits the file base and offset\n+\n+ Returns: (filename, filebase, offset)\n+\n+ If no file is identified, None is returned\n+ '''\n+ fname = self.getFileByVa(va)\n+ if not fname:\n+ return None\n+\n+ fbase = self.getFileMeta(fname, 'imagebase')\n+ off = va - fbase\n+ return (fname, fbase, off)\n+\ndef addLocation(self, va, size, ltype, tinfo=None):\n\"\"\"\nAdd a location tuple.\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/cli.py",
"new_path": "vivisect/cli.py",
"diff": "@@ -184,7 +184,12 @@ class VivCli(e_cli.EnviCli, vivisect.VivWorkspace):\nregex = re.compile(line, re.I)\nfor va, name in self.getNames():\nif regex.search(name):\n- self.vprint('0x%.8x: %s' % (va, name))\n+ ftup = self.getFileAndOffset(va)\n+ if ftup:\n+ fname, fbase, off = ftup\n+ self.vprint('0x%.8x: %s (%r + 0x%x)' % (va, name, fname, off))\n+ else:\n+ self.vprint(\"WARNING: 0x%x: %r doesn't exist in a file!\" % (va, name))\ndef do_save(self, line):\n\"\"\"\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -310,6 +310,9 @@ class VivisectTest(v_t_utils.VivTest):\nself.chgrp_vw.do_names('plt_.*64')\noutput = self.chgrp_vw.canvas.strval\nself.assertIn(\"0x08048e60: chgrp.plt_fseeko64\", output)\n+ self.chgrp_vw.do_names('__entry')\n+ output = self.chgrp_vw.canvas.strval\n+ self.assertIn(\"0x080491e0: chgrp.__entry ('chgrp' + 0x11e0)\", output)\nself.chgrp_vw.canvas.clearCanvas()\ndef test_cli_pathcount(self):\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
improved output from the do_names command. (#516)
|
718,770 |
27.04.2022 07:31:06
| 14,400 |
162fffdd911ff5475f7f80d1b111db6ea12a72eb
|
somehow got this one wrong.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/vivbin.py",
"new_path": "vivisect/vivbin.py",
"diff": "@@ -7,8 +7,8 @@ import argparse\nimport cProfile\nimport importlib.util\n+import envi.exc as e_exc\nimport envi.common as e_common\n-import envi.config as e_config\nimport envi.threads as e_threads\nimport vivisect.cli as viv_cli\n@@ -64,7 +64,7 @@ def main():\ntry:\nvw.config.parseConfigOption(option)\n- except e_config.ConfigNoAssignment as e:\n+ except e_exc.ConfigNoAssignment as e:\nlogger.critical(vw.config.reprConfigPaths() + \"\\n\")\nlogger.critical(e)\nlogger.critical(\"syntax: \\t-O <secname>.<optname>=<optval> (optval must be json syntax)\")\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
somehow got this one wrong. (#518)
|
718,765 |
28.04.2022 12:35:31
| 14,400 |
f42c3ca5bfc15404f0894659ffd1ab1df6b1018c
|
libc_start_main logging fix
Fix a logging message in libc_start_main
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/libc_start_main.py",
"new_path": "vivisect/analysis/elf/libc_start_main.py",
"diff": "@@ -87,7 +87,7 @@ class AnalysisMonitor(viv_imp_monitor.AnalysisMonitor):\nif op.iflags & envi.IF_CALL:\n# it's a call, get the target\nbranches = [br for br in op.getBranches(emu) if not (br[1] & envi.BR_FALL)]\n- logger.debug('libc_start_main: 0x%x\\ttgts: %r', self.startmain, branches)\n+ logger.debug('libc_start_main: %r\\t\\ttgts: %r', [hex(x) for x in self.startmain], branches)\n# check if it matches what we believe to be __libc_start_main\nfor branch in branches:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
libc_start_main logging fix (#519)
Fix a logging message in libc_start_main
|
718,765 |
28.04.2022 19:12:43
| 14,400 |
7ef22b78aace366599fc6c143a93ca193d533ce0
|
V1.0.8 Release Prep
V1.0.7 Release Prep
|
[
{
"change_type": "MODIFY",
"old_path": ".bumpversion.cfg",
"new_path": ".bumpversion.cfg",
"diff": "[bumpversion]\n-current_version = 1.0.7\n+current_version = 1.0.8\n[bumpversion:file:setup.py]\nsearch = VERSION = '{current_version}'\n"
},
{
"change_type": "MODIFY",
"old_path": "CHANGELOG.rst",
"new_path": "CHANGELOG.rst",
"diff": "Vivisect Changelog\n******************\n+V1.0.8 - 2022-04-28\n+===================\n+\n+Features\n+========\n+- Improved Save-As capabilities when connected to a remote server and better struct making from the UI.\n+ (`#501 <https://github.com/vivisect/vivisect/pull/501>`_)\n+- Improve output for the UI's ``names`` command.\n+ (`#516 <https://github.com/vivisect/vivisect/pull/516>`_)\n+\n+Fixes\n+=====\n+- Fix issue in the proxy case where we forgot to snap in the analysis modules.\n+ (`#498 <https://github.com/vivisect/vivisect/pull/498>`_)\n+- Fix string naming.\n+ (`#502 <https://github.com/vivisect/vivisect/pull/502>`_)\n+- Fix a bug in ELFPLT analysis where certain dynamic tables were missing.\n+ (`#503 <https://github.com/vivisect/vivisect/pull/503>`_)\n+- Fix an issue where ELF parsing of STT_FUNCs was based on too many bits.\n+ (`#505 <https://github.com/vivisect/vivisect/pull/505>`_)\n+- Fix an missing name issue in Save-As.\n+ (`#507 <https://github.com/vivisect/vivisect/pull/507>`_)\n+- Improve thread safety for client workspaces.\n+ (`#508 <https://github.com/vivisect/vivisect/pull/508>`_)\n+- Fix the i386 Emulator's handling of rep(n)z.\n+ (`#513 <https://github.com/vivisect/vivisect/pull/513>`_)\n+- Fix issue when dealing with invalid PE section names.\n+ (`#514 <https://github.com/vivisect/vivisect/pull/514>`_)\n+- Fix an incorrect import name in vivbin.\n+ (`#518 <https://github.com/vivisect/vivisect/pull/518>`_)\n+- Fix a debug logging message in the ``libc_start_main`` analysis pass that would cause that analysis pass to exception out.\n+ (`#519 <https://github.com/vivisect/vivisect/pull/519>`_)\n+\nV1.0.7 - 2022-01-13\n===================\n@@ -338,4 +371,4 @@ Fixes\nv0.1.0rc1 - 2020-07-30\n======================\n-- Initial Pypi Release\n+- Initial PyPI Release\n"
},
{
"change_type": "MODIFY",
"old_path": "setup.py",
"new_path": "setup.py",
"diff": "from setuptools import find_packages, setup\nfrom os import path\n-VERSION = '1.0.7'\n+VERSION = '1.0.8'\ndirn = path.abspath(path.dirname(__file__))\nwith open(path.join(dirn, 'README.md'), 'r') as fd:\n"
},
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -2299,6 +2299,6 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these yourself\n-version = (1, 0, 7)\n+version = (1, 0, 8)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -3193,6 +3193,6 @@ def getVivPath(*pathents):\n##############################################################################\n# The following are touched during the release process by bump2version.\n# You should have no reason to modify these directly\n-version = (1, 0, 7)\n+version = (1, 0, 8)\nverstring = '.'.join([str(x) for x in version])\ncommit = ''\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
V1.0.8 Release Prep (#520)
V1.0.7 Release Prep
|
718,770 |
09.05.2022 14:27:55
| 14,400 |
aa7c3e9290846bb333b1ff1044ec112657a00a08
|
update impapi to cover msvcr100.dll
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/impapi/windows/i386.py",
"new_path": "vivisect/impapi/windows/i386.py",
"diff": "@@ -7325,3 +7325,6 @@ api = {\n'mfc42.?afxbeginthread@@ygpavcwinthread@@p6aipax@z0hikpau_security_attributes@@@z':( 'int', None, 'cdecl', 'mfc42.?AfxBeginThread@@YGPAVCWinThread@@P6AIPAX@Z0HIKPAU_SECURITY_ATTRIBUTES@@@Z', (('void *','funcptr'), ('void *','ptr'), ('int',None), ('int',None), ('int',None), ('void *','ptr')) ),\n} # END\n+\n+msvcr100 = {'msvcr100.' + name.split('.', 1)[1] : data for name, data in api.items() if name.startswith('msvcrt.')}\n+\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
update impapi to cover msvcr100.dll (#522)
|
718,770 |
11.05.2022 13:33:35
| 14,400 |
94f4e65aa842f23d89acfadd2a15c59f60c6f029
|
one last thing.
funcrgraph bugfix (for backwards compat)
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/remote/server.py",
"new_path": "vivisect/remote/server.py",
"diff": "@@ -205,7 +205,7 @@ class VivServer:\nlock, fpath, pevents, users, leaders, leaderloc = wsinfo\noldclient = False\n- elif chan in self._req_wsinfo:\n+ elif chan in self.wsdict:\n# DEPRECATED: this is for backwards compat. use only the chandict code one year from today, 5/10/2022.\nwsname = chan\nwsinfo = self._req_wsinfo(wsname)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
one last thing. (#523)
funcrgraph bugfix (for backwards compat)
|
718,770 |
16.05.2022 22:36:31
| 14,400 |
c17653195d5e03ec4257a7a10629e08f529426ef
|
Easy Bugfix: update SaveToServer dialog
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/qt/remote.py",
"new_path": "vivisect/qt/remote.py",
"diff": "@@ -98,7 +98,20 @@ class VivSaveServerDialog(VivServerDialog):\ndef __init__(self, vw, parent=None):\nVivServerDialog.__init__(self, vw, parent=parent)\n+\nself.wsname = QLineEdit(vw.getMeta('StorageName', ''), parent=self)\n+ wslabel = QLabel(\"Storage Path:\")\n+\n+ bot_box = QWidget(parent=self)\n+ hbox = QHBoxLayout(bot_box)\n+ hbox.setContentsMargins(2, 2, 2, 2)\n+ hbox.setSpacing(4)\n+ hbox.addWidget(wslabel)\n+ hbox.addWidget(self.wsname)\n+\n+ layout = self.layout()\n+ layout.insertRow(1, bot_box)\n+\nself.setWindowTitle('Save to Workspace Server...')\ndef getNameAndServer(self):\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Easy Bugfix: update SaveToServer dialog (#527)
|
718,770 |
17.05.2022 15:00:05
| 14,400 |
ec9032d396e13926001f51fdb84e5f50bb504f0e
|
log loadPeIntoWorkspace and args
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/pe.py",
"new_path": "vivisect/parsers/pe.py",
"diff": "@@ -83,6 +83,10 @@ relmap = {\ndef loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\n+ if baseaddr:\n+ logger.info(\"loadPeIntoWorkspace(%r, %r, baseaddr=0x%x\", pe, filename, baseaddr)\n+ else:\n+ logger.info(\"loadPeIntoWorkspace(%r, %r, baseaddr=%r\", pe, filename, baseaddr)\nmach = pe.IMAGE_NT_HEADERS.FileHeader.Machine\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
log loadPeIntoWorkspace and args
|
718,776 |
25.05.2022 07:18:00
| -7,200 |
b8565c5019a28676c0e0d4389712ec9088efcbfa
|
Update windows.py
Add `Ex` to imphooks
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/impemu/platarch/windows.py",
"new_path": "vivisect/impemu/platarch/windows.py",
"diff": "@@ -80,14 +80,14 @@ class WindowsMixin(v_i_emulator.WorkspaceEmulator):\ndef kernel32_LoadLibraryExW(self, emu, callconv, api, argv):\nreturn self.kernel32_LoadLibraryW(emu, callconv, api, argv)\n- @imphook('kernel32.GetModuleHandleA')\n+ @imphook('kernel32.GetModuleHandleExA')\ndef kernel32_GetModuleHandleExA(self, emu, callconv, api, argv):\ndwFlags,lpLibName,phModule = argv\nlibname = self.readLibraryPath(lpLibName, unicode=False)\nretval = emu.setVivTaint('dynlib',libname)\ncallconv.execCallReturn(emu, retval, len(argv))\n- @imphook('kernel32.GetModuleHandleW')\n+ @imphook('kernel32.GetModuleHandleExW')\ndef kernel32_GetModuleHandleExA(self, emu, callconv, api, argv):\ndwFlags,lpLibName,phModule = argv\nlibname = self.readLibraryPath(lpLibName, unicode=True)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Update windows.py (#530)
Add `Ex` to imphooks
|
718,770 |
17.06.2022 18:08:22
| 14,400 |
5048cb7675fcb2f83e05993e052dd77179253ec3
|
move the "file not based at 0" decision to the individual parsers. blob certainly cares.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2794,8 +2794,6 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\n# get baseaddr and size, then make sure we have a good baseaddr\nbaseaddr, size = mod.getMemBaseAndSize(self, filename=filename, baseaddr=baseaddr)\nlogger.debug('initial baseva: 0x%x size: 0x%x', baseaddr, size)\n- if baseaddr == 0:\n- baseaddr = 0x300000\nbaseaddr = self.findFreeMemoryBlock(size, baseaddr)\nlogger.debug(\"loading %r (size: 0x%x) at 0x%x\", filename, size, baseaddr)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/defconfig.py",
"new_path": "vivisect/defconfig.py",
"diff": "@@ -9,11 +9,13 @@ defconfig = {\n'parsers':{\n'pe':{\n+ 'baseaddr': 0x200000,\n'loadresources':False,\n'carvepes':True,\n'nx':False,\n},\n'elf':{\n+ 'baseaddr': 0x200000,\n},\n'blob':{\n'arch':'',\n@@ -60,11 +62,13 @@ docconfig = {\n'parsers':{\n'pe':{\n+ 'baseaddr': 'Address used to relocate PE files if base-address is 0 and PE is relocatable',\n'loadresources':'Should we load resource segments?',\n'carvepes':'Should we carve pes?',\n'nx':'Should we truly treat sections that dont execute as non executable?'\n},\n'elf':{\n+ 'baseaddr': 'Address used to relocate ELF files if base-address is 0 and ELF is relocatable',\n},\n'blob':{\n'arch':'What architecture is the blob?',\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -64,6 +64,10 @@ def getMemBaseAndSize(vw, filename, baseaddr=None):\ntopmem = endva\nsize = topmem - baseaddr\n+\n+ if baseaddr == 0:\n+ baseaddr = vw.config.viv.parsers.elf.baseoffset\n+\nif savebase:\n# if we provided a baseaddr, override what the file wants\nbaseaddr = savebase\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/pe.py",
"new_path": "vivisect/parsers/pe.py",
"diff": "@@ -610,6 +610,10 @@ def getMemBaseAndSize(vw, filename, baseaddr=None):\ntopmem = endva\nsize = topmem - baseaddr\n+\n+ if baseaddr == 0:\n+ baseaddr = vw.config.viv.parsers.pe.baseoffset\n+\nif savebase:\n# if we provided a baseaddr, override what the file wants\nbaseaddr = savebase\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
move the "file not based at 0" decision to the individual parsers. blob certainly cares.
|
718,765 |
22.07.2022 00:11:07
| 14,400 |
1d079e01b10ab6c9756ef27f13d8109bee71d0d2
|
Docbuild Fixes
|
[
{
"change_type": "MODIFY",
"old_path": ".readthedocs.yaml",
"new_path": ".readthedocs.yaml",
"diff": "+version: 2\n+\npython:\nversion: 3.8\n- pip_install: true\n- extra_requirements:\n- - docs\n+ install:\n+ - requirements: requirements_doc.txt\n"
},
{
"change_type": "MODIFY",
"old_path": "CHANGELOG.rst",
"new_path": "CHANGELOG.rst",
"diff": "@@ -6,14 +6,14 @@ V1.0.8 - 2022-04-28\n===================\nFeatures\n-========\n+--------\n- Improved Save-As capabilities when connected to a remote server and better struct making from the UI.\n(`#501 <https://github.com/vivisect/vivisect/pull/501>`_)\n- Improve output for the UI's ``names`` command.\n(`#516 <https://github.com/vivisect/vivisect/pull/516>`_)\nFixes\n-=====\n+-----\n- Fix issue in the proxy case where we forgot to snap in the analysis modules.\n(`#498 <https://github.com/vivisect/vivisect/pull/498>`_)\n- Fix string naming.\n"
},
{
"change_type": "MODIFY",
"old_path": "README.md",
"new_path": "README.md",
"diff": "@@ -31,3 +31,5 @@ be loadable in python3 vivisect.\n## Build Status\n[](https://circleci.com/gh/vivisect/vivisect/tree/master)\n+\n+[](https://vivisect.readthedocs.io/en/latest/?badge=latest)\n"
},
{
"change_type": "MODIFY",
"old_path": "requirements_doc.txt",
"new_path": "requirements_doc.txt",
"diff": "-r requirements_dev.txt\nsphinx>=1.8.2,<2.0.0\nsphinx-rtd-theme>=0.4.2,<1.0.0\n+jinja2<3.1.0\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Docbuild Fixes (#535)
|
718,770 |
08.08.2022 08:20:34
| 14,400 |
e2a204fb75479e9f04f3caa198a47cf8afedae68
|
Update vivisect/analysis/elf/elfplt.py
pre checking both FPLT and FPLTSZ for None individually
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/elf/elfplt.py",
"new_path": "vivisect/analysis/elf/elfplt.py",
"diff": "@@ -162,7 +162,7 @@ def getPLTs(vw):\nFPLTSZ = fdyns.get('DT_PLTRELSZ')\n# if we don't have FPLT or FPLTSZ, skip this\n- if None in (FPLT, FPLTSZ):\n+ if FPLT is None or FPLTSZ is None:\ncontinue\nif vw.getFileMeta(fname, 'addbase'):\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Update vivisect/analysis/elf/elfplt.py
pre @rakuy0, checking both FPLT and FPLTSZ for None individually
|
718,770 |
01.09.2022 09:41:18
| 14,400 |
848338daaa8814e85de355b189a93b6d24f79640
|
add api entry for *.__read_chk on posix
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/impapi/posix/i386.py",
"new_path": "vivisect/impapi/posix/i386.py",
"diff": "@@ -1118,6 +1118,7 @@ api = {\n'*.readdir_r': ('int', None, 'cdecl', '*.readdir_r', (('DIR *', 'dirstream'), ('struct dirent *', 'entry'), ('struct dirent * *', 'result'))),\n'*.readlink': ('ssize_t', None, 'cdecl', '*.readlink', (('const char *', 'filename'), ('char *', 'buffer'), ('size_t', 'size'))),\n'*.readv': ('ssize_t', None, 'cdecl', '*.readv', (('int', 'filedes'), ('const struct iovec *', 'vector'), ('int', 'count'))),\n+ '*.__read_chk': ('ssize_t', None, 'cdecl', '*.__read_chk', (('int', 'filedes'), ('void *', 'buffer'), ('size_t', 'nbytes'), ('size_t', 'buflen'))),\n'*.realloc': ('void *', None, 'cdecl', '*.realloc', (('void *', 'ptr'), ('size_t', 'newsize'))),\n'*.reallocarray': ('void *', None, 'cdecl', '*.reallocarray', (('void *', 'ptr'), ('size_t', 'nmemb'), ('size_t', 'size'))),\n'*.realpath': ('char *', None, 'cdecl', '*.realpath', (('const char * restrict', 'name'), ('char * restrict', 'resolved'))),\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
add api entry for *.__read_chk on posix (#545)
|
718,770 |
01.09.2022 11:25:36
| 14,400 |
96e70b5f74e80dfdd2d2a6f62a89bf2a1f5f3f7b
|
fix opcache key creation
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -1167,7 +1167,7 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nif loctup is not None and loctup[L_TINFO] and loctup[L_LTYPE] == LOC_OP:\narch = loctup[L_TINFO]\nif not skipcache:\n- key = (va, arch, b[:16])\n+ key = (va, arch, b[off:off+16])\nvalu = self._op_cache.get(key, None)\nif not valu:\nvalu = self.imem_archs[(arch & envi.ARCH_MASK) >> 16].archParseOpcode(b, off, va)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
fix opcache key creation (#544)
|
718,770 |
09.09.2022 14:22:06
| 14,400 |
d11eb54be2bc682ffc85ef2cbe2a5c5b6d7c3e65
|
pe bugfix and start-of-cleanup
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/pe.py",
"new_path": "vivisect/parsers/pe.py",
"diff": "@@ -333,7 +333,7 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\nif sec.SizeOfRawData > pe.filesize:\ncontinue\n- plen = sec.VirtualSize - sec.SizeOfRawData\n+ #plen = sec.VirtualSize - sec.SizeOfRawData\ntry:\n# According to http://code.google.com/p/corkami/wiki/PE#section_table if SizeOfRawData is larger than VirtualSize, VS is used..\n@@ -589,7 +589,7 @@ def getMemBaseAndSize(vw, filename, baseaddr=None):\nnbase = nsec.VirtualAddress + baseaddr\nmlen = nbase - secbase\n- memmaps.append((secbase, 7, '', plen))\n+ memmaps.append((secbase, 7, '', mlen))\ncontinue\nif sec.SizeOfRawData < sec.VirtualSize:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
pe bugfix and start-of-cleanup
|
718,770 |
14.09.2022 08:15:12
| 14,400 |
1eb944f130aaa46ea6d547a03eccccdc0e2b2f9e
|
macho log-fix per
change to blob file to fix bug in file lookups
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/blob.py",
"new_path": "vivisect/parsers/blob.py",
"diff": "@@ -36,9 +36,9 @@ def parseFd(vw, fd, filename=None, baseaddr=None):\nvw.setMeta('DefaultCall', archcalls.get(arch, 'unknown'))\nbytez = fd.read()\n- vw.addMemoryMap(baseaddr, 7, filename, bytez)\n- vw.addSegment(baseaddr, len(bytez), '%.8x' % baseaddr, 'blob')\nfname = vw.addFile(filename, baseaddr, v_parsers.md5Bytes(bytez))\n+ vw.addMemoryMap(baseaddr, 7, fname, bytez)\n+ vw.addSegment(baseaddr, len(bytez), '%.8x' % baseaddr, fname)\nvw.setFileMeta(fname, 'sha256', v_parsers.sha256Bytes(bytez))\ndef parseFile(vw, filename, baseaddr=None):\n@@ -64,8 +64,8 @@ def parseFile(vw, filename, baseaddr=None):\nbytez = f.read()\nfname = vw.addFile(filename, baseaddr, v_parsers.md5File(filename))\nvw.setFileMeta(fname, 'sha256', v_parsers.sha256Bytes(bytez))\n- vw.addMemoryMap(baseaddr, 7, filename, bytez)\n- vw.addSegment(baseaddr, len(bytez), '%.8x' % baseaddr, 'blob')\n+ vw.addMemoryMap(baseaddr, 7, fname, bytez)\n+ vw.addSegment(baseaddr, len(bytez), '%.8x' % baseaddr, fname)\ndef parseMemory(vw, memobj, baseaddr):\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/macho.py",
"new_path": "vivisect/parsers/macho.py",
"diff": "@@ -151,7 +151,7 @@ def _loadMacho(vw, filebytes, filename=None, baseaddr=None):\nlogger.debug(\"Applying struct %r at 0x%x\", sname, va+offset)\nvw.makeStructure(va + offset, sname)\nexcept Exception as e:\n- print(\"Error: %r\" % e)\n+ logger.warning(\"Error: %r\", e)\nfor va, cmdname in cmdinfo:\nvw.makeName(va, \"macho_\" + cmdname)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
macho log-fix per @rakuy0
change to blob file to fix bug in file lookups
|
718,770 |
14.09.2022 09:41:12
| 14,400 |
7609a9bb2e09ef8ade69aaa8f7bdc77a13b226ca
|
admin rights stuff per (and a try/finally clause around getAdminRights())
moving connectImportsWithExports() into the linker analysis module instead of the VivWorkspace
remove superfluous `op = vw.parseOpcode()` call
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2858,64 +2858,6 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\noff = va - fileva\nself._fireEvent(VWE_WRITEMEM, (fname, off, bytez, self._supervisor))\n- def connectImportsWithExports(self):\n- \"\"\"\n- Look for any \"imported\" symbols that are satisfied by current exports.\n- Wire up the connection.\n-\n- Currently this is simply a pointer write at the location of the Import.\n- If this behavior is ever insufficient, we'll want to track the special\n- nature through the Export/Import events.\n- \"\"\"\n- logger.info('linking Imports with Exports')\n- # store old setting and set _supervisor mode (so we can write wherever\n- # we want, regardless of permissions)\n- oldsup = self._supervisor\n- self._supervisor = True\n-\n- for iva, isz, itype, isym in self.getImports():\n- impfname, impsym = isym.split('.', 1)\n- for eva, num, esym, efname in self.getExports():\n- if impsym != esym:\n- continue\n-\n- if impfname not in (efname, '*'):\n- continue\n-\n- if self.isFunctionThunk(eva):\n- logger.info(\"Skipping Exported Thunk\")\n- continue\n-\n- # file and symbol name match. apply the magic.\n- # do we ever *not* write in the full address at the import site?\n- logger.debug(\"connecting Import 0x%x -> Export 0x%x (%r)\", iva, eva, isym)\n- self.writeMemoryPtr(iva, eva)\n-\n- # remove the LOC_IMPORT and make it a Pointer instead\n- self.delLocation(iva)\n- self.makePointer(iva, follow=False) # don't follow, it'll be analyzed later?\n-\n- # store the former Import in a VaSet\n- self.setVaSetRow('ResolvedImports', (iva, isym, eva))\n-\n- # check if any xrefs to the import are branches and make code-xrefs for them\n- for xrfr, xrto, xrt, xrflags in self.getXrefsTo(iva):\n- loc = self.getLocation(xrfr)\n- if not loc:\n- continue\n-\n- lva, lsz, ltype, ltinfo = loc\n- if ltype != LOC_OP:\n- logger.warning(\"XREF not from an Opcode: 0x%x -> 0x%x (%r)\", lva, eva, loc)\n- continue\n-\n- op = self.parseOpcode(lva)\n- self.addXref(lva, eva, REF_CODE)\n- logger.debug(\"addXref(0x%x -> 0x%x)\", lva, eva)\n-\n- # restore previous supervisor mode\n- self._supervisor = oldsup\n-\ndef getFiles(self):\n\"\"\"\nReturn the current list of file objects in this\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/generic/linker.py",
"new_path": "vivisect/analysis/generic/linker.py",
"diff": "Connect any Exports we have to Imports we may also have\n(most useful for multiple file workspaces)\n'''\n+\n+import logging\n+logger = logging.getLogger(__name__)\n+\n+\n+def connectImportsWithExports(vw):\n+ \"\"\"\n+ Look for any \"imported\" symbols that are satisfied by current exports.\n+ Wire up the connection.\n+\n+ Currently this is simply a pointer write at the location of the Import.\n+ If this behavior is ever insufficient, we'll want to track the special\n+ nature through the Export/Import events.\n+ \"\"\"\n+ logger.info('linking Imports with Exports')\n+ # store old setting and set _supervisor mode (so we can write wherever\n+ # we want, regardless of permissions)\n+ with vw.getAdminRights():\n+ for iva, isz, itype, isym in vw.getImports():\n+ impfname, impsym = isym.split('.', 1)\n+ for eva, num, esym, efname in vw.getExports():\n+ if impsym != esym:\n+ continue\n+\n+ if impfname not in (efname, '*'):\n+ continue\n+\n+ if vw.isFunctionThunk(eva):\n+ logger.info(\"Skipping Exported Thunk\")\n+ continue\n+\n+ # file and symbol name match. apply the magic.\n+ # do we ever *not* write in the full address at the import site?\n+ logger.debug(\"connecting Import 0x%x -> Export 0x%x (%r)\", iva, eva, isym)\n+ vw.writeMemoryPtr(iva, eva)\n+\n+ # remove the LOC_IMPORT and make it a Pointer instead\n+ vw.delLocation(iva)\n+ vw.makePointer(iva, follow=False) # don't follow, it'll be analyzed later?\n+\n+ # store the former Import in a VaSet\n+ vw.setVaSetRow('ResolvedImports', (iva, isym, eva))\n+\n+ # check if any xrefs to the import are branches and make code-xrefs for them\n+ for xrfr, xrto, xrt, xrflags in vw.getXrefsTo(iva):\n+ loc = vw.getLocation(xrfr)\n+ if not loc:\n+ continue\n+\n+ lva, lsz, ltype, ltinfo = loc\n+ if ltype != LOC_OP:\n+ logger.warning(\"XREF not from an Opcode: 0x%x -> 0x%x (%r)\", lva, eva, loc)\n+ continue\n+\n+ vw.addXref(lva, eva, REF_CODE)\n+ logger.debug(\"addXref(0x%x -> 0x%x)\", lva, eva)\n+\n+\ndef analyze(vw):\n- vw.connectImportsWithExports()\n+ connectImportsWithExports(vw)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/base.py",
"new_path": "vivisect/base.py",
"diff": "@@ -210,9 +210,20 @@ class VivWorkspaceCore(viv_impapi.ImportApi):\n@contextlib.contextmanager\ndef getAdminRights(self):\n+ '''\n+ Support function for ContextManager to support usage like:\n+ \"with vw.getAdminWrites():\"\n+\n+ Sets _supervisor privileges (allowing writes to all workspace maps),\n+ yields to perform the user's write-action, then sets _supervisor back\n+ to the original value.\n+ '''\n+ oldrights = self._supervisor\nself._supervisor = True\n+ try:\nyield\n- self._supervisor = False\n+ finally:\n+ self._supervisor = oldrights\ndef _handleADDLOCATION(self, loc):\nlva, lsize, ltype, linfo = loc\n@@ -544,10 +555,8 @@ class VivWorkspaceCore(viv_impapi.ImportApi):\n'''\nfname, off, bytez, supv = einfo\nimgbase = self.getFileMeta(fname, 'imagebase')\n- savesupv = self._supervisor\n- self._supervisor = True\n+ with self.getAdminRights():\ne_mem.MemoryObject.writeMemory(self, imgbase + off, bytez)\n- self._supervisor = savesupv\ndef _initEventHandlers(self):\nself.ehand = [None for x in range(VWE_MAX)]\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
admin rights stuff per @rakuy0 (and a try/finally clause around getAdminRights())
moving connectImportsWithExports() into the linker analysis module instead of the VivWorkspace
remove superfluous `op = vw.parseOpcode()` call
|
718,770 |
14.09.2022 12:05:53
| 14,400 |
9ded4ca0e1f62436185172d0ecfe712599d489a9
|
avoid double-parsing: part 1 (ELF loads without apparent flaws)
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2807,13 +2807,6 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nmod = viv_parsers.getParserModule(fmtname)\n- # get baseaddr and size, then make sure we have a good baseaddr\n- baseaddr, size = mod.getMemBaseAndSize(self, filename=filename, baseaddr=baseaddr)\n- logger.debug('initial baseva: 0x%x size: 0x%x', baseaddr, size)\n-\n- baseaddr = self.findFreeMemoryBlock(size, baseaddr)\n- logger.debug(\"loading %r (size: 0x%x) at 0x%x\", filename, size, baseaddr)\n-\nfname = mod.parseFile(self, filename=filename, baseaddr=baseaddr)\nif not self.getMeta('StorageName'):\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -44,15 +44,12 @@ def parseFd(vw, fd, filename=None, baseaddr=None):\ndef parseMemory(vw, memobj, baseaddr):\nraise Exception('FIXME implement parseMemory for elf!')\n-def getMemBaseAndSize(vw, filename, baseaddr=None):\n+def getMemBaseAndSize(vw, elf, baseaddr=None):\n'''\nReturns the default baseaddr and memory size required to load the file\n'''\nsavebase = baseaddr\n- fd = open(filename, 'rb')\n- elf = Elf.Elf(fd)\n-\nmemmaps = getMemoryMapInfo(elf)\nbaseaddr = 0xffffffffffffffffffffffff\ntopmem = 0\n@@ -260,6 +257,15 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nnew_pointers = []\nnew_functions = []\n+ # get baseaddr and size, then make sure we have a good baseaddr\n+ filebaseaddr, size = getMemBaseAndSize(vw, elf, baseaddr=baseaddr)\n+ if baseaddr is None:\n+ baseaddr = filebaseaddr\n+\n+ logger.debug('initial file baseva: 0x%x size: 0x%x', baseaddr, size)\n+ baseaddr = vw.findFreeMemoryBlock(size, baseaddr)\n+ logger.debug(\"loading %r (size: 0x%x) at 0x%x\", filename, size, baseaddr)\n+\narch = arch_names.get(elf.e_machine)\nif arch is None:\nraise Exception(\"Unsupported Architecture: %d\\n\", elf.e_machine)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
avoid double-parsing: part 1 (ELF loads without apparent flaws)
|
718,770 |
18.09.2022 20:04:36
| 14,400 |
d496c8f9ca344cf516e4687f466e2f00b8e71bf2
|
most things are initially roughed in and tested.
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/ihex.py",
"new_path": "vivisect/parsers/ihex.py",
"diff": "@@ -67,23 +67,12 @@ def parseFile(vw, filename, baseaddr=None):\ndef parseMemory(vw, memobj, baseaddr):\nraise Exception('ihex loader cannot parse memory!')\n-def getMemBaseAndSize(vw, filename, baseaddr=None):\n+def getMemBaseAndSize(vw, ihex, baseaddr=None):\n'''\nReturns the default baseaddr and memory size required to load the file\n'''\nsavebase = baseaddr\n- # figure out of there's an offset into the file we need to skip\n- offset = vw.config.viv.parsers.ihex.offset\n- if not offset:\n- offset = 0\n- ihex = v_ihex.IHexFile()\n- with open(filename, 'rb') as f:\n- shdr = f.read(offset)\n- sbytes = f.read()\n-\n- ihex.vsParse(sbytes)\n-\nmemmaps = ihex.getMemoryMaps()\nbaseaddr = 0xffffffffffffffffffffffff\ntopmem = 0\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/macho.py",
"new_path": "vivisect/parsers/macho.py",
"diff": "@@ -84,31 +84,41 @@ def checkFatMagicAndTrunc(vw, filebytes):\ndef _loadMacho(vw, filebytes, filename=None, baseaddr=None):\n+ hdr = e_common.hexify(filebytes[:4])\n+\n+ # Check for the FAT binary magic...\n+ if hdr in ('cafebabe', 'bebafeca'):\n+ filebytes = checkFatMagicAndTrunc(vw, filebytes)\n+\n+ # Instantiate the parser wrapper and parse bytes\n+ macho = vs_macho.mach_o()\n+ macho.vsParse(filebytes)\n+ # find the lowest loadable address, then get an offset so we can apply it later\n+ fakebase, size = getMemBaseAndSize(vw, macho, baseaddr=baseaddr))\n+ offset = baseaddr - fakebase\n+ logger.debug('initial file baseva: 0x%x size: 0x%x (address offset: 0x%x)', fakebase, size, offset)\n+\n+ # Determine base address to load into\n# We fake them to *much* higher than norm so pointer tests do better...\n+ # 1) baseaddr arg\n+ # 2) vw.config.viv.parsers.macho.baseaddr if non-zero\n+ # 3) vw.findFreeMemoryBlock() if baseaddr is zero/None or already exists in another map (collision)\n+\nif baseaddr is None:\nbaseaddr = vw.config.viv.parsers.macho.baseaddr\n+ if vw.isValidPointer(baseaddr):\n+ logger.info(\"baseaddr (0x%x) is in use. Finding appropriate address base.\", baseaddr)\n+ baseaddr = vw.findFreeMemoryBlock(size, fakebase)\n+ logger.debug(\"loading %r (size: 0x%x) at 0x%x\", filename, size, baseaddr)\n+\nif filename is None:\nfilename = 'macho_%.8x' % baseaddr # FIXME more than one!\n- # find the lowest loadable address, then get an offset so we can apply it later\n- fakebase, size = getMemBaseAndSize(vw, filename)\n- offset = baseaddr - fakebase\n- logger.debug(\"address offset: 0x%x\" % offset)\n-\n# grab md5 and sha256 hashes before we modify the bytes\nfhash = viv_parsers.md5Bytes(filebytes)\nsha256 = viv_parsers.sha256Bytes(filebytes)\n- hdr = e_common.hexify(filebytes[:4])\n-\n- # Check for the FAT binary magic...\n- if hdr in ('cafebabe', 'bebafeca'):\n- filebytes = checkFatMagicAndTrunc(vw, filebytes)\n-\n- # Instantiate the parser wrapper and parse bytes\n- macho = vs_macho.mach_o()\n- macho.vsParse(filebytes)\narch = vs_macho.mach_cpu_names.get(macho.mach_header.cputype)\nif arch is None:\n@@ -168,19 +178,13 @@ def _loadMacho(vw, filebytes, filename=None, baseaddr=None):\nreturn fname\n-def getMemBaseAndSize(vw, filename, baseaddr=None):\n+def getMemBaseAndSize(vw, macho, baseaddr=None):\n'''\nReturns the default baseaddr and memory size required to load the file\n'''\nif baseaddr is None:\nbaseaddr = vw.config.viv.parsers.macho.baseaddr\n- with open(filename, 'rb') as f:\n- filebytes = f.read()\n-\n- macho = vs_macho.mach_o()\n- macho.vsParse(filebytes)\n-\nmemmaps = macho.getSegments()\nbaseaddr = 0xffffffffffffffffffffffff\ntopmem = 0\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/pe.py",
"new_path": "vivisect/parsers/pe.py",
"diff": "@@ -83,10 +83,14 @@ relmap = {\ndef loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\n- if baseaddr:\n- logger.info(\"loadPeIntoWorkspace(%r, %r, baseaddr=0x%x\", pe, filename, baseaddr)\n- else:\n- logger.info(\"loadPeIntoWorkspace(%r, %r, baseaddr=%r\", pe, filename, baseaddr)\n+ # get baseaddr and size, then make sure we have a good baseaddr\n+ filebaseaddr, size = getMemBaseAndSize(vw, pe, baseaddr=baseaddr)\n+ if baseaddr is None:\n+ baseaddr = filebaseaddr\n+\n+ logger.debug('initial file baseva: 0x%x size: 0x%x', baseaddr, size)\n+ baseaddr = vw.findFreeMemoryBlock(size, baseaddr)\n+ logger.info(\"loadPeIntoWorkspace: loading %r (size: 0x%x) at 0x%x\", filename, size, baseaddr)\nmach = pe.IMAGE_NT_HEADERS.FileHeader.Machine\n@@ -560,14 +564,12 @@ def loadPeIntoWorkspace(vw, pe, filename=None, baseaddr=None):\nreturn fname\n-def getMemBaseAndSize(vw, filename, baseaddr=None):\n+def getMemBaseAndSize(vw, pe, baseaddr=None):\n'''\nReturns the default baseaddr and memory size required to load the file\n'''\nsavebase = baseaddr\n- pe = PE.PE(open(filename, 'rb'))\n-\nif baseaddr is None:\nbaseaddr = pe.IMAGE_NT_HEADERS.OptionalHeader.ImageBase\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/srec.py",
"new_path": "vivisect/parsers/srec.py",
"diff": "@@ -65,23 +65,12 @@ def parseFile(vw, filename, baseaddr=None):\ndef parseMemory(vw, memobj, baseaddr):\nraise Exception('srec loader cannot parse memory!')\n-def getMemBaseAndSize(vw, filename, baseaddr=None):\n+def getMemBaseAndSize(vw, srec, baseaddr=None):\n'''\nReturns the default baseaddr and memory size required to load the file\n'''\nsavebase = baseaddr\n- # figure out of there's an offset into the file we need to skip\n- offset = vw.config.viv.parsers.srec.offset\n- if not offset:\n- offset = 0\n- srec = v_srec.SRecFile()\n- with open(filename, 'rb') as f:\n- shdr = f.read(offset)\n- sbytes = f.read()\n-\n- srec.vsParse(sbytes)\n-\nmemmaps = srec.getMemoryMaps()\nbaseaddr = 0xffffffffffffffffffffffff\ntopmem = 0\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/teststorage.py",
"new_path": "vivisect/tests/teststorage.py",
"diff": "@@ -65,8 +65,8 @@ class StorageTests(unittest.TestCase):\nold = list(vw.exportWorkspace())\nnew = list(ovw.exportWorkspace())\n- self.assertEqual(len(old), 38)\n- self.assertEqual(len(new), 39) # the last event is a setMeta made by loadWorkspace\n+ self.assertEqual(len(old), 39)\n+ self.assertEqual(len(new), 40) # the last event is a setMeta made by loadWorkspace\nself.assertEqual(new[-1], (VWE_SETMETA, ('StorageName', self.tmpf.name)))\nfor idx in range(len(old)):\nself.assertEqual(old[idx], new[idx])\n@@ -96,7 +96,7 @@ class StorageTests(unittest.TestCase):\nmvw._event_list = []\nmvw.loadWorkspace(mpfile.name)\nmevt = list(mvw.exportWorkspace())\n- self.assertEqual(len(mevt), 39)\n+ self.assertEqual(len(mevt), 40)\nbvw = vivisect.VivWorkspace()\nbvw.setMeta('StorageModule', 'vivisect.storage.basicfile')\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
most things are initially roughed in and tested.
|
718,770 |
30.09.2022 08:14:07
| 14,400 |
c74322be800d53bc60cf3f1ca5a0f77eb1cf34c3
|
unittest tweaks and minor bugfix
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/macho.py",
"new_path": "vivisect/parsers/macho.py",
"diff": "@@ -94,7 +94,7 @@ def _loadMacho(vw, filebytes, filename=None, baseaddr=None):\nmacho = vs_macho.mach_o()\nmacho.vsParse(filebytes)\n# find the lowest loadable address, then get an offset so we can apply it later\n- fakebase, size = getMemBaseAndSize(vw, macho, baseaddr=baseaddr))\n+ fakebase, size = getMemBaseAndSize(vw, macho, baseaddr=baseaddr)\noffset = baseaddr - fakebase\nlogger.debug('initial file baseva: 0x%x size: 0x%x (address offset: 0x%x)', fakebase, size, offset)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/teststorage.py",
"new_path": "vivisect/tests/teststorage.py",
"diff": "@@ -103,7 +103,7 @@ class StorageTests(unittest.TestCase):\nbvw._event_list = []\nbvw.loadWorkspace(basicfile.name)\nbevt = list(bvw.exportWorkspace())\n- self.assertEqual(len(bevt), 39)\n+ self.assertEqual(len(bevt), 40)\n# the last three events are specific to the different storage modules\nfor idx in range(len(mevt) - 3):\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect_multi.py",
"new_path": "vivisect/tests/testvivisect_multi.py",
"diff": "@@ -81,7 +81,7 @@ class VivisectMultiFileTest(v_t_utils.VivTest):\nself.assertEqual(chgrp_main-chgrp, 5712)\n# test Relocs\n- self.assertIn(('ld_2_31', 179948, 2, 135749), vw.getRelocations())\n+ self.assertIn(('ld_2_31', 179948, 2, 135749, 4), vw.getRelocations())\nrel0 = vw.parseExpression('ld_2_31 + 179948')\nptr0 = vw.parseExpression('ld_2_31 + 135749')\nself.assertIn((rel0, ptr0, 3, 0), vw.getXrefsFrom(rel0))\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
unittest tweaks and minor bugfix
|
718,770 |
29.10.2022 16:38:59
| 14,400 |
8e9bba09f3867aa09e7aabc68480b00b50f968e4
|
Enabling POSIX Library Load notifications (since PTRACE event model doesn't alert on library loads like Windows does).
a couple other general bugfixes/deprecation-cleanup "because we're in the area."
(this is ripped out of the Vtrace PR because it's important enough to get it's own)
|
[
{
"change_type": "MODIFY",
"old_path": "envi/threads.py",
"new_path": "envi/threads.py",
"diff": "@@ -19,8 +19,7 @@ def firethread(func):\nand callers may not expect sync behavior!\n'''\ndef dothread(*args, **kwargs):\n- thr = threading.Thread(target=func, args=args, kwargs=kwargs)\n- thr.setDaemon(True)\n+ thr = threading.Thread(target=func, args=args, kwargs=kwargs, daemon=True)\nthr.start()\nreturn thr\nfunctools.update_wrapper(dothread, func)\n@@ -44,8 +43,7 @@ def maintthread(stime):\ntime.sleep(stime)\ndef dothread(*args, **kwargs):\n- thr = threading.Thread(target=maintloop, args=args, kwargs=kwargs)\n- thr.setDaemon(True)\n+ thr = threading.Thread(target=maintloop, args=args, kwargs=kwargs, daemon=True)\nthr.start()\nfunctools.update_wrapper(dothread, func)\n"
},
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -492,6 +492,7 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\ntrace.setMeta('PendingBreak', False)\nbp = trace.getCurrentBreakpoint()\nif bp:\n+ if not bp.silent:\nself.vprint(\"Thread: %d Hit Break: %s\" % (tid, repr(bp)))\ncmdstr = self.bpcmds.get(bp.id, None)\nif cmdstr is not None:\n@@ -574,7 +575,7 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\nif dohex:\nmemstr = binascii.unhexlify(memstr)\nif douni:\n- memstr = (\"\\x00\".join(memstr)) + \"\\x00\"\n+ memstr = (b\"\\x00\".join(memstr)) + b\"\\x00\"\naddr = self.parseExpression(exprstr)\nself.memobj.writeMemory(addr, memstr)\n@@ -1861,6 +1862,8 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\nself.vprint(\" [ Breakpoints ]\")\nfor bp in self.trace.getBreakpoints():\n+ if bp.untouchable: # don't list untouchable bp's\n+ continue\nself._print_bp(bp)\ndef _print_bp(self, bp):\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/__init__.py",
"new_path": "vtrace/__init__.py",
"diff": "@@ -156,6 +156,11 @@ class Trace(e_mem.IMemory, e_reg.RegisterContext, e_resolv.SymbolResolver, objec\n# Add event numbers to here for auto-continue\nself.auto_continue = [NOTIFY_LOAD_LIBRARY, NOTIFY_CREATE_THREAD, NOTIFY_UNLOAD_LIBRARY, NOTIFY_EXIT_THREAD, NOTIFY_DEBUG_PRINT]\n+ # Create a LoadLibrary hook to enable simple and consistent\n+ # Break-On-Load/Init functionality. This is also necessary for\n+ # resolving symbols on new library loads.\n+ self.registerNotifier(NOTIFY_LOAD_LIBRARY, LibraryNotifier())\n+\ndef execute(self, cmdline):\n\"\"\"\nStart a new process and debug it\n@@ -694,7 +699,7 @@ class Trace(e_mem.IMemory, e_reg.RegisterContext, e_resolv.SymbolResolver, objec\n\"\"\"\nself.requireAttached()\nbp = self.bpbyid.pop(id, None)\n- if bp is not None:\n+ if bp is not None and not bp.untouchable:\nbp.deactivate(self)\nif bp in self.deferred:\nself.deferred.remove(bp)\n@@ -708,6 +713,20 @@ class Trace(e_mem.IMemory, e_reg.RegisterContext, e_resolv.SymbolResolver, objec\n# Remove cached breakpoint code\nBreakpoint.bpcodeobj.pop(id, None)\n+ def _updateBreakAddresses(self):\n+ \"\"\"\n+ Update breakpoint address resolution (in unresolved breakpoints).\n+ Intended to be run after events which change the namespace, such as\n+ NOTIFY_LOAD_LIBRARY events\n+ \"\"\"\n+ for bp in self.deferred:\n+ bp.resolveAddress(self)\n+ if bp.address is not None:\n+ self.breakpoints[bp.address] = bp\n+ self.deferred.remove(bp)\n+ bp.activate(self)\n+ logger.warning(\"Resolved bp address: %r\", bp)\n+\ndef getCurrentBreakpoint(self):\n\"\"\"\nReturn the current breakpoint otherwise None\n@@ -755,7 +774,7 @@ class Trace(e_mem.IMemory, e_reg.RegisterContext, e_resolv.SymbolResolver, objec\nNOTE: code which wants to be remote-safe should use this\n\"\"\"\nbp = self.getBreakpoint(bpid)\n- if bp is None:\n+ if bp is None or bp.untouchable:\nraise Exception(\"Breakpoint %d Not Found\" % bpid)\nif not enabled: # To catch the \"disable\" of fastbreaks...\nbp.deactivate(self)\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/breakpoints.py",
"new_path": "vtrace/breakpoints.py",
"diff": "@@ -31,6 +31,9 @@ class Breakpoint:\nself.active = False\nself.fastbreak = False # no NOTIFY_BREAK, autocont, no NOTIFY_CONTINUE\nself.stealthbreak = False # no NOTIFY_BREAK\n+ self.silent = False\n+ self.untouchable = False # system breakpoint. can't remove it or see it.\n+ self._complained = False\nself.id = -1\nself.vte = None\n@@ -96,9 +99,13 @@ class Breakpoint:\nif self.address is None and self.vte:\ntry:\nself.address = trace.parseExpression(self.vte)\n+\nexcept Exception as e:\n- logger.warning('Failed to resolve breakpoint address for expression: %s', self.vte)\n- logger.warning('Error:', exc_info=1)\n+ # this will happen with unresolved breakpoints.\n+ # depending on when resolution happens, the library may not have loaded yet.\n+ if not self._complained:\n+ logger.warning('Failed to resolve breakpoint address for expression: %s (delayed resolution?)', self.vte)\n+ self._complained = True\nself.address = None\n# If we resolved, lets get our saved code...\n@@ -404,3 +411,17 @@ class PostHookBreakpoint(NiceBreakpoint):\nret_addr, args = tup\nself.runPostHookCallbacks(event, trace, ret_addr, args)\n+\n+class PosixLibLoadHookBreakpoint(Breakpoint):\n+ '''\n+ POSIX systems need to hook DL to identfy when libraries are loaded.\n+ '''\n+ def __init__(self, expression):\n+ Breakpoint.__init__(self, None, expression=expression)\n+ self.untouchable = True\n+ self.silent = True\n+\n+ def notify(self, event, trace):\n+ logger.debug(\"PosixLibLoadHookBreakpoint: reanalyze maps and resolve symbols\")\n+ trace._findLibraryMaps(b'\\x7fELF', always=True)\n+ # handle unresolved expression bp's\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/notifiers.py",
"new_path": "vtrace/notifiers.py",
"diff": "@@ -136,3 +136,12 @@ class DistributedNotifier(Notifier):\ndef deregisterNotifier(self, event, notif):\nnlist = self.notifiers.get(event)\nnlist.remove(notif)\n+\n+class LibraryNotifier(Notifier):\n+ def notify(self, event, trace):\n+ logger.info(\"LibraryNotifier.notify(%r, %r)\", event, trace)\n+\n+ # update unresolved breakpoints:\n+ trace._updateBreakAddresses()\n+\n+\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/platforms/base.py",
"new_path": "vtrace/platforms/base.py",
"diff": "@@ -537,6 +537,7 @@ class TracerBase(vtrace.Notifier):\nabout a LOAD_LIBRARY. (This means *not* from inside another\nnotifer)\n\"\"\"\n+ logger.info(\"addLibraryBase(%r, 0x%x, %r)\", libname, address, always)\nself.setMeta(\"LatestLibrary\", None)\nself.setMeta(\"LatestLibraryNorm\", None)\n@@ -878,7 +879,7 @@ class TracerBase(vtrace.Notifier):\ndef threadwrap(func):\ndef trfunc(self, *args, **kwargs):\n- if threading.currentThread().__class__ == TracerThread:\n+ if threading.current_thread().__class__ == TracerThread:\nreturn func(self, *args, **kwargs)\n# Proxy the call through a single thread\nq = queue.Queue()\n@@ -904,9 +905,8 @@ class TracerThread(threading.Thread):\nto make particular calls and on what platforms... YAY!\n\"\"\"\ndef __init__(self):\n- threading.Thread.__init__(self)\n+ threading.Thread.__init__(self, daemon=True)\nself.queue = queue.Queue()\n- self.setDaemon(True)\nself.start()\ndef run(self):\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/platforms/linux.py",
"new_path": "vtrace/platforms/linux.py",
"diff": "@@ -413,6 +413,12 @@ class LinuxMixin(v_posix.PtraceMixin, v_posix.PosixMixin):\nraise Exception(\"PT_ATTACH failed!\")\nself.setMeta(\"ExeName\", self._findExe(pid))\n+ def _LibraryLoadHack(self):\n+ # drop special breakpoint at ld._dl_catch_exception\n+ import vtrace.breakpoints as v_bp\n+ bp = v_bp.PosixLibLoadHookBreakpoint('ld._dl_catch_exception')\n+ self.addBreakpoint(bp)\n+\ndef platformPs(self):\npslist = []\nfor dname in self.platformListDir('/proc'):\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/platforms/posix.py",
"new_path": "vtrace/platforms/posix.py",
"diff": "@@ -61,6 +61,14 @@ class PosixMixin:\n# break after our library load events to make things easy\nself.runAgain(False) # Clear this, if they want BREAK to run, it will\nself.fireNotifiers(vtrace.NOTIFY_BREAK)\n+ # POSIX hack - Windows signals on library load, POSIX doesn't\n+ self._LibraryLoadHack()\n+\n+ def _LibraryLoadHack(self):\n+ '''\n+ Implement at the platform level\n+ '''\n+ pass\ndef platformProcessEvent(self, event):\npid, status = event\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/tests/testbasic.py",
"new_path": "vtrace/tests/testbasic.py",
"diff": "@@ -59,6 +59,6 @@ class VtraceBasicTest(vt_tests.VtraceProcessTest):\nclass VtraceBasicExecTest(VtraceBasicTest, vt_tests.VtraceExecTest):\nbreakpoints = {\n'windows': 'ntdll.NtTerminateProcess',\n- 'linux': 'ld.malloc',\n+ 'linux': 'libc.malloc',\n'freebsd': 'ld._rtld_thread_init',\n}\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Enabling POSIX Library Load notifications (since PTRACE event model doesn't alert on library loads like Windows does).
a couple other general bugfixes/deprecation-cleanup "because we're in the area."
(this is ripped out of the Vtrace PR because it's important enough to get it's own)
|
718,770 |
30.10.2022 19:17:48
| 14,400 |
70269ab567e9ebc333faf79db11e961e4d4fdcbc
|
Update vstruct/defs/macho/__init__.py
|
[
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/macho/__init__.py",
"new_path": "vstruct/defs/macho/__init__.py",
"diff": "@@ -54,7 +54,7 @@ class mach_o(vstruct.VStruct):\nself._entrypoints = []\noffset = len(self.mach_header)\nfor fname, vs in self.load_commands:\n- if vs.cmd == LC_UNIXTHREAD:\n+ if vs.cmd in (LC_MAIN, LC_UNIXTHREAD):\neoff = len(vs)\npsize = vs.flavor\nbigend = self.vsGetEndian()\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Update vstruct/defs/macho/__init__.py
|
718,770 |
19.11.2022 14:23:38
| 18,000 |
063c9d2071b3a0ecdc9697edb4556208dbca5cb0
|
fix i386's vtrace archGetBackTrace() results (they're reversed from what the rest of the system do)
|
[
{
"change_type": "MODIFY",
"old_path": "vtrace/archs/i386.py",
"new_path": "vtrace/archs/i386.py",
"diff": "@@ -144,7 +144,7 @@ class i386Mixin(e_i386.i386Module, e_i386.i386RegisterContext, i386WatchMixin):\nebp, eip = struct.unpack(\"<LL\", buf)\nif frames[-1] == (ebp, eip):\nbreak\n- frames.append((ebp, eip))\n+ frames.append((eip, ebp))\ncurrent += 1\nexcept:\nbreak\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
fix i386's vtrace archGetBackTrace() results (they're reversed from what the rest of the system do) (#553)
|
718,770 |
19.11.2022 14:49:00
| 18,000 |
fb23b9f2ba86ff7f36975bb088367ddca7c64887
|
minor bugfixes: VDB RegisterView widget and MSP430 archGetRegisterGroups
|
[
{
"change_type": "MODIFY",
"old_path": "envi/archs/msp430/__init__.py",
"new_path": "envi/archs/msp430/__init__.py",
"diff": "@@ -24,11 +24,9 @@ class Msp430Module(envi.ArchitectureModule):\ndef archGetNopInstr(self):\nreturn b'\\x03\\x43' # NOP is emulated with: MOV #0, R3\n- def archGetRegisterGroups(self):\n- groups = envi.ArchitectureModule.archGetRegisterGroups(self)\n- general= ('general', registers, )\n- groups.append(general)\n- return groups\n+ def initRegGroups(self):\n+ envi.ArchitectureModule.initRegGroups(self)\n+ self._regGrps.update({'general': registers})\ndef getPointerSize(self):\nreturn 2\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/qt.py",
"new_path": "vtrace/qt.py",
"diff": "@@ -183,7 +183,7 @@ class RegistersView(QWidget):\nself.flagviews = {}\nreg_groups = trace.archGetRegisterGroups()\n- for name, group in reg_groups:\n+ for name, group in reg_groups.items():\nself.regviews[name] = group\nself.viewnames.addItem(name)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
minor bugfixes: VDB RegisterView widget and MSP430 archGetRegisterGroups (#552)
Co-authored-by: James Gross <45212823+rakuy0@users.noreply.github.com>
|
718,770 |
21.11.2022 00:35:07
| 18,000 |
8d508dc3cfcb7f0d6e216667fdb7a228d60713b4
|
elf bugfixes (test)
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -117,7 +117,7 @@ def getMemoryMapInfo(elf, fname=None, baseaddr=None):\nbaseaddr = 0x05000000\nfor offset,size in merged:\nbytez = elf.readAtOffset(offset,size)\n- memmaps.append((baseaddr + offset, 0x7, fname, bytez, None))\n+ memmaps.append((baseaddr + offset, 0x7, fname, bytez, 0))\nfor sec in secs:\nif sec.sh_offset and sec.sh_size:\n@@ -229,6 +229,11 @@ def getAddBaseAddr(elf, baseaddr=None):\nif not elf.isPreLinked() and elf.isSharedObject():\naddbase = True\nbaseoff = baseaddr\n+\n+ elif elf.isRelocatable():\n+ addbase = True\n+ baseoff = 0\n+\nelse:\naddbase = False\nbaseoff = baseaddr - elfbaseaddr\n@@ -301,6 +306,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\n# Base addr is earliest section address rounded to pagesize\n# Some ELF's require adding the baseaddr to most/all later addresses\naddbase, baseoff, baseaddr = getAddBaseAddr(elf, baseaddr)\n+ logger.warning(\"addbase: %r, baseoff: 0x%x, baseaddr: 0x%x\", addbase, baseoff, baseaddr)\nelf.fd.seek(0)\nmd5hash = v_parsers.md5Bytes(byts)\n@@ -421,7 +427,9 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\ncontinue # Skip non-memory mapped sections\nsva = sec.sh_addr\n+ logger.warning(\"sva = sec.sh_addr (0x%x)\", sva)\nsva += baseoff\n+ logger.warning(\"sva += baseoff (0x%x) == 0x%x\", baseoff, sva)\n# if we've already defined a location at this address, skip it. (eg. DYNAMICS)\nif vw.getLocation(sva) == sva:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
elf bugfixes (test)
|
718,770 |
21.11.2022 07:35:13
| 18,000 |
a2dc2ce4c37f5bb75dea1c4e4245b97215efd504
|
remove default baseaddr for ELF's with no Program Headers
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -98,6 +98,7 @@ def getMemoryMapInfo(elf, fname=None, baseaddr=None):\nlogger.info('Skipping: %s', pgm)\nif len(pgms) == 0:\n+ # NOTE: among other file types, Linux Kernel modules have no program headers\nsecs = elf.getSections()\n# fall back to loading sections as best we can...\nlogger.info('elf: no program headers found! (in %r)', fname)\n@@ -114,7 +115,6 @@ def getMemoryMapInfo(elf, fname=None, baseaddr=None):\nmerged.append( maps[i] )\n- baseaddr = 0x05000000\nfor offset,size in merged:\nbytez = elf.readAtOffset(offset,size)\nmemmaps.append((baseaddr + offset, 0x7, fname, bytez, 0))\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
remove default baseaddr for ELF's with no Program Headers
|
718,770 |
22.11.2022 12:04:09
| 18,000 |
06d3c1cf5f715830473f2be5f6b266d6820637be
|
upstream bugfix: alignment issues for Section-only ELFs
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -56,7 +56,7 @@ def getMemBaseAndSize(vw, elf, baseaddr=None):\nfor mapva, mperms, mname, mbytes, malign in memmaps:\nif mapva < baseaddr:\n- baseaddr = mapva\n+ baseaddr = mapva & -e_const.PAGE_SIZE # align to page-size\nendva = mapva + len(mbytes)\nif endva > topmem:\ntopmem = endva\n@@ -103,7 +103,7 @@ def getMemoryMapInfo(elf, fname=None, baseaddr=None):\n# fall back to loading sections as best we can...\nlogger.info('elf: no program headers found! (in %r)', fname)\n- maps = [ [s.sh_offset,s.sh_size] for s in secs if s.sh_offset and s.sh_size ]\n+ maps = [ [s.sh_offset,s.sh_size,s.sh_addralign] for s in secs if s.sh_offset and s.sh_size ]\nmaps.sort()\nmerged = []\n@@ -115,9 +115,9 @@ def getMemoryMapInfo(elf, fname=None, baseaddr=None):\nmerged.append( maps[i] )\n- for offset,size in merged:\n+ for offset,size,align in merged:\nbytez = elf.readAtOffset(offset,size)\n- memmaps.append((baseaddr + offset, 0x7, fname, bytez, 0))\n+ memmaps.append((baseaddr + offset, 0x7, fname, bytez, align))\nfor sec in secs:\nif sec.sh_offset and sec.sh_size:\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
upstream bugfix: alignment issues for Section-only ELFs
|
718,770 |
27.11.2022 19:13:32
| 18,000 |
7c4c2055ae43d1806efa475e5b015127f9ea9f34
|
modify hellokernel unittest to allow for relocation.
add comment to ELF "isRelocatable()" function to describe what this means
|
[
{
"change_type": "MODIFY",
"old_path": "Elf/__init__.py",
"new_path": "Elf/__init__.py",
"diff": "@@ -757,6 +757,8 @@ class Elf(vs_elf.Elf32, vs_elf.Elf64):\ndef isRelocatable(self):\n'''\nReturns true if the given Elf binary is marked as a relocatable file.\n+ isRelocatable() helps determine if this ELF is a Kernel Module (.ko)\n+ or Object file (.o), *not* a Shared Object (.so) or executable.\n'''\nreturn self.e_type == ET_REL\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -1593,7 +1593,7 @@ class VivisectTest(v_t_utils.VivTest):\ndef test_relocatable_elf_simple(self):\nkmod = helpers.getTestWorkspace('linux', 'amd64', 'hellokernel.ko')\n- self.eq(set(kmod.getFunctions()), set([0x7c, 0x95]))\n+ self.eq(set(kmod.getFunctions()), set([0x20007c, 0x200095]))\ndef test_guid(self):\nvw = vivisect.VivWorkspace()\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
modify hellokernel unittest to allow for relocation.
add comment to ELF "isRelocatable()" function to describe what this means
|
718,770 |
29.11.2022 08:14:58
| 18,000 |
1c6b0b2a70998496a2dba1d42648a364f07d8797
|
restore kernel module basing behavior to match master
|
[
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -63,7 +63,7 @@ def getMemBaseAndSize(vw, elf, baseaddr=None):\nsize = topmem - baseaddr\n- if baseaddr == 0:\n+ if baseaddr == 0 and not elf.isRelocatable():\nbaseaddr = vw.config.viv.parsers.elf.baseaddr\nif savebase:\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -1593,7 +1593,7 @@ class VivisectTest(v_t_utils.VivTest):\ndef test_relocatable_elf_simple(self):\nkmod = helpers.getTestWorkspace('linux', 'amd64', 'hellokernel.ko')\n- self.eq(set(kmod.getFunctions()), set([0x20007c, 0x200095]))\n+ self.eq(set(kmod.getFunctions()), set([0x7c, 0x95]))\ndef test_guid(self):\nvw = vivisect.VivWorkspace()\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
restore kernel module basing behavior to match master
|
718,770 |
04.12.2022 19:30:13
| 18,000 |
c7d9190c70585824920ca2963c3c3e9854a5a2fd
|
missed one (per
|
[
{
"change_type": "MODIFY",
"old_path": "vtrace/platforms/posix.py",
"new_path": "vtrace/platforms/posix.py",
"diff": "@@ -62,9 +62,9 @@ class PosixMixin:\nself.runAgain(False) # Clear this, if they want BREAK to run, it will\nself.fireNotifiers(vtrace.NOTIFY_BREAK)\n# POSIX hack - Windows signals on library load, POSIX doesn't\n- self._LibraryLoadHack()\n+ self._LibraryLoadHook()\n- def _LibraryLoadHack(self):\n+ def _LibraryLoadHook(self):\n'''\nImplement at the platform level\n'''\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
missed one (per @rakuy0)
|
718,770 |
28.12.2022 16:38:10
| 18,000 |
7c3bbd2bfdfce519cc53b139758371c704710fb1
|
all the changes and i discussed. (i think)
|
[
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -492,6 +492,7 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\ntrace.setMeta('PendingBreak', False)\nbp = trace.getCurrentBreakpoint()\nif bp:\n+ if not self.silent:\nself.vprint(\"Thread: %d Hit Break: %s\" % (tid, repr(bp)))\ncmdstr = self.bpcmds.get(bp.id, None)\nif cmdstr is not None:\n@@ -1721,7 +1722,7 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\nwhere options include:\n-F toggles a breakpoint's FastBreak mode\n- -S toggles a breakpoint's Stealth mode (doesn't fire Notifiers, still runs BP code)\n+ -S toggles a breakpoint's Silent mode (doesn't print console msg, still runs BP code)\n-V prints more metadata about a given breakpoint (default is just the code)\nNOTE: Your code must be surrounded by \"s and may not\n@@ -1750,7 +1751,7 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\nverbose = True\nelif opt == '-S':\n- bp.stealthbreak = not bp.stealthbreak\n+ bp.silent = not bp.silent\nelif opt == '-F':\nbp.fastbreak = not bp.fastbreak\n@@ -1764,12 +1765,12 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\nself.vprint(\"%r\" % bp)\nself.vprint(\" code: %s\" % (pystr))\nself.vprint(\" FastBreak: %r\" % bp.fastbreak)\n- self.vprint(\" resonce: %r\" % bp.resonce)\n- self.vprint(\" active: %r\" % bp.active)\nself.vprint(\" enabled: %r\" % bp.enabled)\n+ self.vprint(\" silent: %r\" % bp.silent)\n+ self.vprint(\" active: %r\" % bp.active)\n+ self.vprint(\" resonce: %r\" % bp.resonce)\n+ if bp.stealthbreak:\nself.vprint(\" stealth: %r\" % bp.stealthbreak)\n- if bp.untouchable:\n- self.vprint(\" untouchable: %r\" % bp.untouchable)\nelse:\nself.vprint(\"[%d] Breakpoint code: %s\" % (bpid,pystr))\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/breakpoints.py",
"new_path": "vtrace/breakpoints.py",
"diff": "@@ -25,14 +25,14 @@ class Breakpoint:\nbpcodeobj = {} # Cache compiled code objects on the class def\ndef __init__(self, address, expression=None):\n- self.resonce = False\n+ self.resonce = False # has this addr expression been resolved yet?\nself.address = address\n- self.enabled = True\n- self.active = False\n+ self.enabled = True # should this BP be used/ignored\n+ self.active = False # have we placed a BP in the code (eg. i386: \\xCC)\n+ self.silent = False # don't print \"Hit Break\" messages, still runs code/notifiers\nself.fastbreak = False # no NOTIFY_BREAK, autocont, no NOTIFY_CONTINUE\n- self.stealthbreak = False # no NOTIFY_BREAK\n- self.untouchable = False # system breakpoint. can't remove it or see it.\n- self._complained = False\n+ self.stealthbreak = False # no NOTIFY_BREAK - used for hidden/system events\n+ self._complained = False # only complain about not resolving this *once*\nself.id = -1\nself.vte = None\n@@ -417,10 +417,12 @@ class PosixLibLoadHookBreakpoint(Breakpoint):\n'''\ndef __init__(self, expression):\nBreakpoint.__init__(self, None, expression=expression)\n- self.untouchable = True\nself.stealthbreak = True\ndef notify(self, event, trace):\nlogger.debug(\"PosixLibLoadHookBreakpoint: reanalyze maps and resolve symbols\")\n- trace._findLibraryMaps(b'\\x7fELF', always=True)\n+ if not trace._findLibraryMaps(b'\\x7fELF', always=True):\n+ # if we find new maps, we'll let the LOAD_LIBRARY Autoload config setting handle\n+ # whether we continue or not. if we fire this and *don't* find a new map,\n+ # let's just continue like nothing ever happened. \"Nothing to see here.\"\ntrace.runAgain()\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/platforms/base.py",
"new_path": "vtrace/platforms/base.py",
"diff": "@@ -576,9 +576,11 @@ class TracerBase(vtrace.Notifier):\ndef _findLibraryMaps(self, magic, always=False):\n# A utility for platforms which lack library load\n# notification through the operating system\n+ # TODO: update to handle *losing* memory maps as well.\nbmaps = self.getMeta(\"BadMaps\", [])\ndone = {}\nmlen = len(magic)\n+ newcount = 0\nfor addr, size, perms, fname in self.getMemoryMaps():\nif not fname:\n@@ -587,7 +589,7 @@ class TracerBase(vtrace.Notifier):\nif done.get(fname):\ncontinue\n- if addr in self.getMeta(\"LibraryPaths\"):\n+ if fname == self.getMeta(\"LibraryPaths\").get(addr):\ncontinue\nif fname in bmaps:\n@@ -597,10 +599,13 @@ class TracerBase(vtrace.Notifier):\nif self.readMemory(addr, mlen) == magic:\ndone[fname] = True\nself.addLibraryBase(fname, addr, always=always)\n+ newcount += 1\nexcept Exception as e:\nlogger.warning('findLibraryMaps(0x%x, %d, %s, %s) hit exception: %s', addr, size, perms, fname, e)\n+ return newcount\n+\ndef _loadBinaryNorm(self, normname):\nif not self.libloaded.get(normname, False):\nfname = self.libpaths.get(normname)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
all the changes @invisig0th and i discussed. (i think)
|
718,770 |
28.12.2022 16:48:07
| 18,000 |
d1f6b9c3bab867b7aeeb280cdc08f477e7f09b58
|
untouchable -> stealthbreak
|
[
{
"change_type": "MODIFY",
"old_path": "vdb/__init__.py",
"new_path": "vdb/__init__.py",
"diff": "@@ -1843,7 +1843,7 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\nelif opt == \"-C\":\nfor bp in self.trace.getBreakpoints():\n- if bp.untouchable:\n+ if bp.stealthbreak:\ncontinue\nself.bpcmds.pop(bp.id, None)\n@@ -1914,8 +1914,8 @@ class Vdb(e_cli.EnviMutableCli, v_notif.Notifier, v_util.TraceManager):\nself.vprint(\" [ Breakpoints ]\")\nfor bp in self.trace.getBreakpoints():\n- if bp.untouchable and not showall:\n- # don't list untouchable bp's (unless forced)\n+ if bp.stealthbreak and not showall:\n+ # don't list stealthbreak bp's (unless forced)\ncontinue\nself._print_bp(bp)\n"
},
{
"change_type": "MODIFY",
"old_path": "vtrace/__init__.py",
"new_path": "vtrace/__init__.py",
"diff": "@@ -699,7 +699,7 @@ class Trace(e_mem.IMemory, e_reg.RegisterContext, e_resolv.SymbolResolver, objec\n\"\"\"\nself.requireAttached()\nbp = self.bpbyid.pop(id, None)\n- if bp is not None and not bp.untouchable:\n+ if bp is not None and not bp.stealthbreak:\nbp.deactivate(self)\nif bp in self.deferred:\nself.deferred.remove(bp)\n@@ -760,7 +760,7 @@ class Trace(e_mem.IMemory, e_reg.RegisterContext, e_resolv.SymbolResolver, objec\nNOTE: code which wants to be remote-safe should use this\n\"\"\"\nbp = self.getBreakpoint(bpid)\n- if bp is None or bp.untouchable:\n+ if bp is None or bp.stealthbreak:\nraise Exception(\"Breakpoint %d Not Found\" % bpid)\nif not enabled: # To catch the \"disable\" of fastbreaks...\nbp.deactivate(self)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
untouchable -> stealthbreak
|
718,770 |
30.12.2022 08:46:17
| 18,000 |
4e24cfcac316482f4e5517f838ec843e779afc48
|
final change to completely supplant PR 557
|
[
{
"change_type": "MODIFY",
"old_path": "envi/memory.py",
"new_path": "envi/memory.py",
"diff": "@@ -146,12 +146,15 @@ class IMemory:\nmmap = self.getMemoryMap(va)\nif mmap is None:\nreturn False\n+\nmapva, mapsize, mapperm, mapfile = mmap\nmapend = mapva+mapsize\nif va+size > mapend:\nreturn False\n- if mapperm & perm != perm:\n+\n+ if mapperm & perm != perm and not self._supervisor:\nreturn False\n+\nreturn True\ndef allocateMemory(self, size, perms=MM_RWX, suggestaddr=0):\n@@ -578,7 +581,7 @@ class MemoryObject(IMemory):\nfor mva, mmaxva, mmap, mbytes in self._map_defs:\nif mva <= va < mmaxva:\nmva, msize, mperms, mfname = mmap\n- if not mperms & MM_READ:\n+ if not (mperms & MM_READ or self._supervisor):\nmsg = \"Bad Memory Read (no READ permission): %s: %s\" % (hex(va), hex(size))\nif _origva is not None:\nmsg += \" (original va: %s)\" % hex(_origva)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
final change to completely supplant PR 557
|
718,770 |
30.12.2022 08:48:07
| 18,000 |
7be248c08a88d8cd531daa6f2f1eb65471e942ce
|
minor tweak and a minor bugfix in linker
|
[
{
"change_type": "MODIFY",
"old_path": "Elf/__init__.py",
"new_path": "Elf/__init__.py",
"diff": "@@ -87,6 +87,7 @@ class Elf(vs_elf.Elf32, vs_elf.Elf64):\nelse:\nself.r_types = {}\n+ self.dyns = {}\nself.pheaders = []\nself.sections = []\nself.secnames = {}\n@@ -290,7 +291,6 @@ class Elf(vs_elf.Elf32, vs_elf.Elf64):\nThis must be run before most Dynamic-data accessors like getDynStrtabString(),\ngetDynSymTabInfo(), etc..\n'''\n- self.dyns = {}\ndynbytes = self.getDynBytes()\nif dynbytes is None:\nreturn\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/analysis/generic/linker.py",
"new_path": "vivisect/analysis/generic/linker.py",
"diff": "@@ -5,7 +5,7 @@ Connect any Exports we have to Imports we may also have\nimport logging\nlogger = logging.getLogger(__name__)\n-from vivisect import LOC_OP\n+from vivisect import LOC_OP, REF_CODE\ndef analyze(vw):\n\"\"\"\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -399,7 +399,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nmakeFunctionTable(elf, vw, f_preinita, f_preinitasz, 'preinit_array', new_functions, new_pointers, baseoff)\n# dynamic table\n- phdr = elf.getDynPHdr() # file offset?\n+ phdr = elf.getDynPHdr() # this is the Program Header which points at the DYNAMICS table, not the other way around.\nif phdr is not None:\nsva, size = phdr.p_vaddr, phdr.p_memsz\nsva += baseoff # getDynInfo returns (offset, filesz)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect_multi.py",
"new_path": "vivisect/tests/testvivisect_multi.py",
"diff": "@@ -9,6 +9,14 @@ logger = logging.getLogger(__name__)\nclass VivisectMultiFileTest(v_t_utils.VivTest):\ndef test_multiple_i386(self):\n+ '''\n+ Note: because we're using a non-relocatable executable (vdir.llvm) and relocating it, this\n+ will unfortunately throw a lot of ugly exceptions, because there are hard-coded addresses\n+ in the instructions themselves, which are relocations because the file isn't \"relocatable\"\n+\n+ We only test the things which should work because they are made available for relocation\n+ through the ELF metadata.\n+ '''\nvw = helpers.getTestWorkspace_nocache('linux', 'i386', 'chgrp.llvm')\nvw = helpers.getTestWorkspace_nocache('linux', 'i386', 'vdir.llvm', vw=vw)\nvw = helpers.getTestWorkspace_nocache('linux', 'i386', 'ld-2.31.so', vw=vw)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
minor tweak and a minor bugfix in linker
|
718,770 |
30.12.2022 16:47:25
| 18,000 |
bd92aa527fbd5044ce81e719649f284701d58394
|
cleanup and mods per
|
[
{
"change_type": "MODIFY",
"old_path": "envi/memory.py",
"new_path": "envi/memory.py",
"diff": "@@ -143,6 +143,7 @@ class IMemory:\nExample probeMemory(0x41414141, 20, envi.memory.MM_WRITE)\n(check if the memory for 20 bytes at 0x41414141 is writable)\n\"\"\"\n+ #FIXME: make probeMemory handle cross-map access as well as _supervisor\nmmap = self.getMemoryMap(va)\nif mmap is None:\nreturn False\n@@ -439,7 +440,7 @@ class MemoryObject(IMemory):\ndef getAdminRights(self):\n'''\nSupport function for ContextManager to support usage like:\n- \"with vw.getAdminWrites():\"\n+ \"with vw.getAdminRights():\"\nSets _supervisor privileges (allowing writes to all workspace maps),\nyields to perform the user's write-action, then sets _supervisor back\n@@ -606,7 +607,7 @@ class MemoryObject(IMemory):\ndef _reqProbeMem(self, va, size, perm):\n'''\n- Calls probeMemory() and either returns or raises the appropriate exception\n+ Checks memory map permissions and either returns True or raises the appropriate exception\n'''\nmsg = None\n@@ -633,7 +634,7 @@ class MemoryObject(IMemory):\nmapva, mapsize, mapperm, mapfile = curmap\n# check permissions\n- if mapperm & perm != perm:\n+ if (mapperm & perm != perm) and not self._supervisor:\nif curmap != startmap:\nmsg = \"Bad Memory Access (%r): 0x%x: 0x%x (orig va: 0x%x)\" % (\ngetPermName(perm), mapva, size, va)\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/__init__.py",
"new_path": "vivisect/__init__.py",
"diff": "@@ -2852,6 +2852,7 @@ class VivWorkspace(e_mem.MemoryObject, viv_base.VivWorkspaceCore):\nOverride writeMemory to hook into the Event subsystem.\nStores overwritten data for easy undo.\n'''\n+ self._reqProbeMem(va, len(bytez), e_mem.MM_WRITE)\noldbytes = self.readMemory(va, len(bytez))\nself._fireEvent(VWE_WRITEMEM, (va, bytez, oldbytes))\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/const.py",
"new_path": "vivisect/const.py",
"diff": "@@ -71,7 +71,7 @@ VWE_CHAT = 40 # (username, message)\nVWE_SYMHINT = 41 # (va, idx, hint)\nVWE_AUTOANALFIN = 42 # (starttime, endtime)\n-VWE_WRITEMEM = 43 # (mapva, offset, bytes)\n+VWE_WRITEMEM = 43 # (va, bytes, oldbytes)\nVWE_MAX = 44\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -300,7 +300,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\nvw.addNoReturnApi(\"*.pthread_exit\")\nvw.addNoReturnApi(\"*.longjmp\")\n- # for VivWorkspace, MSB==1, LSB==0... which is the same as True/False\n+ # for VivWorkspace, MSB==1, LSB==0... which is the same as True/False for Big-Endian\nvw.setEndian(elf.getEndian())\n# Base addr is earliest section address rounded to pagesize\n@@ -427,9 +427,7 @@ def loadElfIntoWorkspace(vw, elf, filename=None, baseaddr=None):\ncontinue # Skip non-memory mapped sections\nsva = sec.sh_addr\n- logger.warning(\"sva = sec.sh_addr (0x%x)\", sva)\nsva += baseoff\n- logger.warning(\"sva += baseoff (0x%x) == 0x%x\", baseoff, sva)\n# if we've already defined a location at this address, skip it. (eg. DYNAMICS)\nif vw.getLocation(sva) == sva:\n@@ -1057,3 +1055,37 @@ def isStripped(elf):\nreturn True\nreturn False\n+\n+\n+'''\n+Base Offsets are complicated, so let's break them down a bit.\n+\n+Base Offsets originally refered to whether to add a base address to values to come up with a virtual address.\n+Variations of ELFs that regularly cause problems/differences:\n+ * Shared Objects\n+ * Kernel Modules / Object files\n+ * Prelinked SO's and EXEs\n+\n+Accesses which are impacted by these addresses/offsets:\n+ * Pointer to Program Headers\n+ * Pointer to Section Headers\n+ * Program Headers\n+ * Sections\n+ * Dynamics Table\n+ * Dynamics Entries (see dt_rebase for existing list which require adding base offsets)\n+ Via either Sections or Dynamics:\n+ * String Table(s)\n+ * Symbol Table(s)\n+ * Relocs Table(s)\n+ * GOT\n+ * PLT\n+ * INIT / FINI\n+ * INIT_ARRAY / FINI_ARRAY / PREINIT_ARRAY\n+ * ??? Dwarf?\n+\n+If a file isn't intended for relocation, this can cause problems because:\n+ * Pointers can be hard-coded into the instructions themselves\n+ * ELF headers/Dynamics can not understand the relocatable parts\n+\n+\n+'''\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -5,6 +5,7 @@ import tempfile\nimport unittest\nimport envi\n+import envi.exc as e_exc\nimport envi.memory as e_memory\nimport envi.memcanvas as e_mcanvas\n@@ -1610,3 +1611,28 @@ class VivisectTest(v_t_utils.VivTest):\n# since it's assigned, the result from \"vw.getVivGuid()\" should be the same\nself.assertEqual(newguid, vw.getVivGuid())\n+\n+ def test_write_fail(self):\n+ with self.snap(self.chown_vw) as vw:\n+ base = vw.getFileMeta('chown', 'imagebase')\n+\n+ oldmem = vw.readMemory(base, 10)\n+\n+ #import envi.interactive as ei; ei.dbg_interact(locals(), globals())\n+ with self.assertRaises(e_exc.SegmentationViolation):\n+ vw.writeMemory(base, b\"testing...\")\n+\n+ self.assertEqual(oldmem, vw.readMemory(base, 10))\n+\n+ with vw.getAdminRights():\n+ vw.writeMemory(base, b\"testing...\")\n+\n+ self.assertEqual(b'testing...', vw.readMemory(base, 10))\n+\n+ with self.assertRaises(e_exc.SegmentationViolation):\n+ vw.writeMemory(base, b\"FOOBARBAZ.\")\n+\n+ self.assertEqual(b'testing...', vw.readMemory(base, 10))\n+\n+ with vw.getAdminRights():\n+ vw.writeMemory(base, oldmem)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
cleanup and mods per @invisig0th
|
718,770 |
30.12.2022 17:15:05
| 18,000 |
d08253f8de49c88201c63eedeb8976cdaf6d28a9
|
_reqProbeMem bugfix!
cleanup isSet->is_set
|
[
{
"change_type": "MODIFY",
"old_path": "envi/cli.py",
"new_path": "envi/cli.py",
"diff": "@@ -259,7 +259,7 @@ class EnviCli(Cmd):\nif intro is not None:\nself.vprint(intro)\n- while not self.shutdown.isSet():\n+ while not self.shutdown.is_set():\ntry:\nCmd.cmdloop(self, intro=intro)\nexcept Exception:\n@@ -311,7 +311,7 @@ class EnviCli(Cmd):\nself.vprint(\"\\nERROR: (%s) %s\" % (msg.__class__.__name__, msg))\nlogger.warning(\"\\nERROR: (%s) %s\", msg.__class__.__name__, msg, exc_info=1)\n- if self.shutdown.isSet():\n+ if self.shutdown.is_set():\nreturn True\ndef do_help(self, line):\n"
},
{
"change_type": "MODIFY",
"old_path": "envi/memory.py",
"new_path": "envi/memory.py",
"diff": "@@ -628,10 +628,10 @@ class MemoryObject(IMemory):\n# check that all memory is valid and correct permissions\nptr = va\ncurmap = startmap\n- endmapva, endmapsize, endmapperm, endmapfile = endmap\n+ endmapva, endmapsz, endmapperm, endmapfile = endmap\nwhile True:\n- mapva, mapsize, mapperm, mapfile = curmap\n+ mapva, mapsz, mapperm, mapfile = curmap\n# check permissions\nif (mapperm & perm != perm) and not self._supervisor:\n@@ -644,11 +644,11 @@ class MemoryObject(IMemory):\nraise envi.SegmentationViolation(va, msg)\n# if the endmap exists, and we are here, and the perms check out... we're good!\n- if ptr == endmapva:\n+ if endmapva <= ptr < endmapva+endmapsz:\nbreak\n# go to next map\n- ptr = mapva + mapsize\n+ ptr = mapva + mapsz\ncurmap = self.getMemoryMap(ptr)\nif not curmap:\nmsg = \"Bad Memory Access (invalid memory range): 0x%x: 0x%x (%s)\" % (\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
_reqProbeMem bugfix!
cleanup isSet->is_set
|
718,766 |
13.01.2023 12:35:53
| 18,000 |
5a3d41efec8c03e0e1d731b55e2f4af327f90670
|
Don't log to root logger
|
[
{
"change_type": "MODIFY",
"old_path": "vstruct/defs/minidump.py",
"new_path": "vstruct/defs/minidump.py",
"diff": "@@ -3,6 +3,8 @@ import logging\nimport vstruct\nfrom vstruct.primitives import *\n+logger = logging.getLogger(__name__)\n+\nclass MiniDumpString(vstruct.VStruct):\ndef __init__(self):\nvstruct.VStruct.__init__(self)\n@@ -570,7 +572,7 @@ class MiniDump(object):\nstream = vars(self)[sclass.__name__] = sclass()\nstream.vsParse(bytez, offset=soffset)\nelse:\n- logging.info('Unknown stream type of %d', header.StreamType)\n+ logger.info('Unknown stream type of %d', header.StreamType)\ndef tree(self):\ntxt = []\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Don't log to root logger (#565)
|
718,772 |
23.01.2023 08:57:30
| 28,800 |
8181858bd4025cc1959edd2482dd0d87bdcd3838
|
Update vamp signatures (replaces old PR#208)
* Update vamp signatures (former PR#208)
* unit test for signatures
* Update vivisect/tests/testvampsigs.py
self.assertEqual instead of assert
* Update vivisect/tests/testvampsigs.py
self.assertEqual instead of assert
* Update vivisect/tests/testvampsigs.py
self.assertEqual instead of assert
|
[
{
"change_type": "ADD",
"old_path": null,
"new_path": "vivisect/tests/testvampsigs.py",
"diff": "+import unittest\n+\n+import vivisect\n+import envi.memory as e_mem\n+\n+\n+class VampSigTests(unittest.TestCase):\n+ '''\n+ Tests to verify detection of signatures for special functions in PE files.\n+ If a signature is detected during makeFunc(), it should mark the\n+ function as a thunk and name it according to the signature.\n+ '''\n+\n+ def run_test(self, opcode_string):\n+ # Create a Vivisect workspace, architecture does not matter.\n+ vw = vivisect.VivWorkspace()\n+ vw.setMeta('Architecture', 'i386')\n+ vw.setMeta('Platform', 'windows')\n+ vw.setMeta('Format', 'pe')\n+\n+ # Add the module that detects signatures.\n+ vw.addFuncAnalysisModule(\"vivisect.analysis.ms.msvc\")\n+\n+ # Put the opcodes into an executable memory map.\n+ mapbase = 0x400000\n+ bufferpgsz = 2 * 4096\n+ vw.addMemoryMap(mapbase - bufferpgsz, e_mem.MM_RWX,\n+ 'test', '@' * bufferpgsz)\n+ bytez = bytes(bytearray.fromhex(opcode_string))\n+ vw.addMemoryMap(mapbase, e_mem.MM_RWX, 'test', bytez)\n+ vw.addSegment(mapbase, len(bytez), 'test_code_%x' % mapbase, 'test')\n+\n+ # Make a function, triggering signature detection.\n+ fva = mapbase\n+ vw.makeFunction(fva)\n+ return vw.getFunctionMeta(fva, 'Thunk')\n+\n+\n+ def test_sig_behavior(self):\n+ '''\n+ Test the behavior of signature detection.\n+ '''\n+ # True signature.\n+ opcodes = '680000000064a10000000050'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.seh3_prolog')\n+\n+ # Incorrect signature.\n+ opcodes = '680000000064a100000000aa'\n+ self.assertEqual(self.run_test(opcodes), None)\n+\n+ # Signature with masked out bytes changed.\n+ opcodes = '68aaaaaaaa64a10000000050'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.seh3_prolog')\n+\n+ # Signature with extra opcodes.\n+ opcodes = '680000000064a100000000504141414141'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.seh3_prolog')\n+\n+ # Signature with opcodes before.\n+ opcodes = '4141680000000064a10000000050'\n+ self.assertEqual(self.run_test(opcodes), None)\n+\n+\n+ def test_seh_sigs(self):\n+ '''\n+ Test detection of seh prolog and epilog signatures.\n+ '''\n+ opcodes = '680000000064a10000000050'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.seh3_prolog')\n+\n+ opcodes = '8b4df064890d00000000595f5e5bc951c3'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.seh3_epilog')\n+\n+ opcodes = '680000000064ff35000000008b442410'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.seh4_prolog')\n+\n+ opcodes = '8b4df064890d00000000595f5f5e5b8be55d51c3'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.seh4_epilog')\n+\n+ opcodes = '8b4df064890d00000000595f5f5e5b8be55d51f2c3'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.seh4_epilog')\n+\n+ opcodes = 'a10000000033c58945fc'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.gs_prolog')\n+\n+\n+ def test_alloca_probe_sigs(self):\n+ '''\n+ Test detection of alloca_probe signatures.\n+ '''\n+ opcodes = '513d001000008d4c2408721481e9001000002d0010000085013d0010000073ec2bc88bc485018be18b088b400450c3'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll._alloca_probe')\n+\n+ opcodes = '518d4c24042bc81bc0f7d023c88bc42500f0ffff3bc8720a8bc159948b00890424c32d001000008500ebe9'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll._alloca_probe')\n+\n+ opcodes = '518d4c24042bc81bc0f7d023c88bc42500f0ffff3bc8f2720b8bc159948b00890424f2c32d001000008500ebe7'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll._alloca_probe')\n+\n+ opcodes = '4883ec104c8914244c895c24084d33db4c8d5424184c2bd04d0f42d3654c8b1c25100000004d3bd37316664181e200f04d8d9b00f0ffff41c603004d3bd375f04c8b14244c8b5c24084883c410c3'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll._alloca_probe')\n+\n+ opcodes = '4883ec104c8914244c895c24084d33db4c8d5424184c2bd04d0f42d3654c8b1c25100000004d3bd3f27317664181e200f04d8d9b00f0ffff41c603004d3bd3f275ef4c8b14244c8b5c24084883c410f2c3'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll._alloca_probe')\n+\n+\n+ def test_security_cookie_sigs(self):\n+ '''\n+ Test detection of security_check_cookie signatures.\n+ '''\n+ opcodes = '3b0d000000007502f3c3e9'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.security_check_cookie')\n+\n+ opcodes = '3b0d00000000f27502f2c3f2e9'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.security_check_cookie')\n+\n+ opcodes = '483b0d00000000751148c1c11066f7c1ffff7502f3c348c1c910e9'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.security_check_cookie_64')\n+\n+ opcodes = '483b0d00000000f2751148c1c11066f7c1fffff27502f2c348c1c910e9'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.security_check_cookie_64')\n+\n+ opcodes = '483b0d00000000f2751248c1c11066f7c1fffff27502f2c348c1c910e9'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.security_check_cookie_64')\n+\n+\n+ def test_gs_failure_sigs(self):\n+ '''\n+ Test detection of gs_failure signatures.\n+ '''\n+ opcodes = '3b0d000000007502f3c3e9'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.security_check_cookie')\n+ opcodes = '8bff558bec5151a300000000890d00000000891500000000891d00000000893500000000893d000000008c15000000008c0d000000008c1d000000008c05000000008c25000000008c2d000000009c'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.report_gsfailure')\n+\n+ opcodes = '8bff558bec81ec28030000a300000000890d00000000891500000000891d00000000893500000000893d00000000668c1500000000668c0d00000000668c1d00000000668c0500000000668c2500000000668c2d000000009c'\n+ self.assertEqual(self.run_test(opcodes), 'ntdll.report_gsfailure')\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/tests/testvivisect.py",
"new_path": "vivisect/tests/testvivisect.py",
"diff": "@@ -1528,7 +1528,8 @@ class VivisectTest(v_t_utils.VivTest):\n0x140049770, 0x140049bf0, 0x140049b80, 0x140049780, 0x140049a00, 0x140049900, 0x140049700,\n0x140049800, 0x140049880, 0x140049c00, 0x140049b00, 0x140049a10, 0x140049790, 0x140049810,\n0x140049a90, 0x140049710, 0x14001ef10, 0x140049890, 0x140049910, 0x140049b90, 0x140049c10,\n- 0x140049b40, 0x140049aa0, 0x1400497a0, 0x140049720, 0x140049820, 0x140049a20, 0x140048a10]\n+ 0x140049b40, 0x140049aa0, 0x1400497a0, 0x140049720, 0x140049820, 0x140049a20, 0x140048a10,\n+ 0x140048a80]\nself.assertEqual(thunks, set(impthunk))\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/vamp/msvc/__init__.py",
"new_path": "vivisect/vamp/msvc/__init__.py",
"diff": "@@ -8,28 +8,45 @@ import envi.bytesig as e_bytesig\nsigs = [\n('680000000064a10000000050', 'ff00000000ffffffffffffff', 'ntdll.seh3_prolog'),\n('8b4df064890d00000000595f5e5bc951c3', None, 'ntdll.seh3_epilog'),\n- ('8b4df064890d00000000595f5f5e5b8be55d51c3', None, 'ntdll.seh4_epilog'),\n('680000000064ff35000000008b442410', 'ff00000000ffffffffffffffffffffff', 'ntdll.seh4_prolog'),\n+ # Seen in 32-bit samples using VS 2005, 2008, 2010, 2012, 2013.\n+ ('8b4df064890d00000000595f5f5e5b8be55d51c3', None, 'ntdll.seh4_epilog'),\n+ # Seen in 32-bit samples using VS 2015, 2017 15.0.\n+ ('8b4df064890d00000000595f5f5e5b8be55d51f2c3', None, 'ntdll.seh4_epilog'),\n('a10000000033c58945fc', 'ff00000000ffffffffff', 'ntdll.gs_prolog'),\n- ('513d001000008d4c2408721481e9001000002d0010000085013d0010000073ec2bc88bc485018be18b088b400450c3', None, 'ntdll._alloca_probe'), #32-bit VS 6.0\n- ('518d4c24042bc81bc0f7d023c88bc42500f0ffff3bc8720a8bc159948b00890424c32d001000008500ebe9', None, 'ntdll._alloca_probe'), # 32-bit VS 2005, 2008, 2010, 2012, 2013\n- ('518d4c24042bc81bc0f7d023c88bc42500f0ffff3bc8f2720b8bc159948b00890424f2c32d001000008500ebe7', None, 'ntdll._alloca_probe'), # 32-bit VS 2015, 2017, 2019\n- ('4883ec104c8914244c895c24084d33db4c8d5424184c2bd04d0f42d3654c8b1c25100000004d3bd37316664181e200f04d8d9b00f0ffff41c603004d3bd375f04c8b14244c8b5c24084883c410c3', None, 'ntdll._alloca_probe'), # 64-bit VS 2005, 2008, 2010, 2012, 2013, 2015\n- ('4883ec104c8914244c895c24084d33db4c8d5424184c2bd04d0f42d3654c8b1c25100000004d3bd3f27317664181e200f04d8d9b00f0ffff41c603004d3bd3f275ef4c8b14244c8b5c24084883c410f2c3', None, 'ntdll._alloca_probe'), # 64-bit VS 2017, 2019\n+ # Seen in 32-bit samples using VS 6.\n+ ('513d001000008d4c2408721481e9001000002d0010000085013d0010000073ec2bc88bc485018be18b088b400450c3', None, 'ntdll._alloca_probe'),\n+ # Seen in 32-bit samples using VS 2005, 2008, 2010, 2012, 2013.\n+ ('518d4c24042bc81bc0f7d023c88bc42500f0ffff3bc8720a8bc159948b00890424c32d001000008500ebe9', None, 'ntdll._alloca_probe'),\n+ # Seen in 32-bit samples using VS 2015, 2017, 2019.\n+ ('518d4c24042bc81bc0f7d023c88bc42500f0ffff3bc8f2720b8bc159948b00890424f2c32d001000008500ebe7', None, 'ntdll._alloca_probe'),\n+ # Seen in 64-bit samples using VS 2005, 2008, 2010, 2012, 2013, 2015.\n+ ('4883ec104c8914244c895c24084d33db4c8d5424184c2bd04d0f42d3654c8b1c25100000004d3bd37316664181e200f04d8d9b00f0ffff41c603004d3bd375f04c8b14244c8b5c24084883c410c3', None, 'ntdll._alloca_probe'),\n+ # Seen in 64-bit samples using VS 2017, 2019.\n+ ('4883ec104c8914244c895c24084d33db4c8d5424184c2bd04d0f42d3654c8b1c25100000004d3bd3f27317664181e200f04d8d9b00f0ffff41c603004d3bd3f275ef4c8b14244c8b5c24084883c410f2c3', None, 'ntdll._alloca_probe'),\n+\n+# ('3b0d000000000f85afdc0200c3', 'ffff00000000ffffffffffffff', 'ntdll.security_check_cookie'),\n+ # 32-bit assembly used in VS 2005, 2008, 2010, 2012, 2013.\n+ ('3b0d000000007502f3c3e9', 'ffff00000000ffffffffff','ntdll.security_check_cookie'),\n+ # 32-bit assembly used in VS 2015, 2017 (includes bnd opcodes).\n+ ('3b0d00000000f27502f2c3f2e9', 'ffff00000000ffffffffffffff', 'ntdll.security_check_cookie'),\n+ # 64-bit assembly used in VS 2005, 2008, 2010, 2012, 2013.\n+ ('483b0d00000000751148c1c11066f7c1ffff7502f3c348c1c910e9', 'ffffff00000000ffffffffffffffffffffffffffffffffffffffff', 'ntdll.security_check_cookie_64'),\n+ # 64-bit assembly used in VS 2015.\n+ ('483b0d00000000f2751148c1c11066f7c1fffff27502f2c348c1c910e9', 'ffffff00000000ffffffffffffffffffffffffffffffffffffffffffff', 'ntdll.security_check_cookie_64'),\n+ # 64-bit assembly used in VS 2019.\n+ ('483b0d00000000f2751248c1c11066f7c1fffff27502f2c348c1c910e9', 'ffffff00000000ffffffffffffffffffffffffffffffffffffffffffff', 'ntdll.security_check_cookie_64'),\n- ('3b0d000000000f85afdc0200c3', 'ffff00000000ffffffffffffff', 'ntdll.security_check_cookie'),\n('8bff558bec5151a300000000890d00000000891500000000891d00000000893500000000893d000000008c15000000008c0d000000008c1d000000008c05000000008c25000000008c2d000000009c',\n'ffffffffffffffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ff',\n'ntdll.report_gsfailure',\n),\n-\n('8bff558bec81ec28030000a300000000890d00000000891500000000891d00000000893500000000893d00000000668c1500000000668c0d00000000668c1d00000000668c0500000000668c2500000000668c2d000000009c',\n'ffffffffffffffffffffffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffff00000000ffffff00000000ffffff00000000ffffff00000000ffffff00000000ffffff00000000ffffff00000000ff',\n'ntdll.report_gsfailure',\n),\n- ('3b0d000000007502f3c3e9', 'ffff00000000ffffffffff','ntdll.security_check_cookie'),\n('6aff5064a100000000508b44240c64892500000000896c240c8d6c240c50c3',None, 'ntdll.eh_prolog'),\n]\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
Update vamp signatures (replaces old PR#208) (#566)
* Update vamp signatures (former PR#208)
* unit test for signatures
* Update vivisect/tests/testvampsigs.py
self.assertEqual instead of assert
* Update vivisect/tests/testvampsigs.py
self.assertEqual instead of assert
* Update vivisect/tests/testvampsigs.py
self.assertEqual instead of assert
Co-authored-by: todd-plantenga <todd.plantenga@mandiant.com>
Co-authored-by: atlas0fd00m <atlas@r4780y.com>
|
718,770 |
02.02.2023 08:57:19
| 18,000 |
21deba74cafcb187b324d74b260e0d89044ece73
|
viv_loader Cleanup and Polish
* IMemory bug discovered by Rakuy0
merge fail in ELF parser (regressing certain related sections)
* minor bugfix for VQVivFuncgraphCanvas.renderMemory
|
[
{
"change_type": "MODIFY",
"old_path": "envi/memory.py",
"new_path": "envi/memory.py",
"diff": "@@ -69,6 +69,7 @@ class IMemory:\nself.setMemArchitecture(arch)\nself.bigend = envi.ENDIAN_LSB\n+ self._supervisor = False\ndef getEndian(self):\n'''\n@@ -434,7 +435,6 @@ class MemoryObject(IMemory):\n\"\"\"\nIMemory.__init__(self, arch=arch)\nself._map_defs = []\n- self._supervisor = False\n@contextlib.contextmanager\ndef getAdminRights(self):\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/parsers/elf.py",
"new_path": "vivisect/parsers/elf.py",
"diff": "@@ -725,7 +725,15 @@ def applyRelocs(elf, vw, addbase=False, baseoff=0):\nfor secidx, r in elf.getRelocs():\nrtype = r.getType()\nrlva = r.r_offset\n+ if elf.isRelocatable():\n+ container = elf.getSectionByIndex(secidx)\n+ if container.sh_flags & elf_lookup.SHF_INFO_LINK:\n+ othr = elf.getSectionByIndex(container.sh_info)\n+ if othr:\n+ rlva += othr.sh_addr\n+\nrlva += baseoff\n+\ntry:\n# If it has a name, it's an externally resolved \"import\" entry,\n# otherwise, just a regular reloc\n"
},
{
"change_type": "MODIFY",
"old_path": "vivisect/qt/funcgraph.py",
"new_path": "vivisect/qt/funcgraph.py",
"diff": "@@ -111,8 +111,14 @@ class VQVivFuncgraphCanvas(vq_memory.VivCanvasBase):\nrunner = functools.partial(self._renderMemoryFinish, cb)\ne_memcanvas.MemoryCanvas.renderMemory(self, va, size, cb=runner)\n- def renderMemory(self, va, size, cb):\n+ def renderMemory(self, va, size, cb=None):\n+ '''\n+ Funcgraph specific renderMemory() function.\n+ '''\n# For the funcgraph canvas, this will be called once per code block\n+ if not cb:\n+ cb = self.__nopcb\n+\nselector = 'codeblock_%.8x' % va\njs = '''var node = document.querySelector(\"#%s\");\nif (node == null) {\n@@ -127,6 +133,9 @@ class VQVivFuncgraphCanvas(vq_memory.VivCanvasBase):\nlogger.log(MIRE, \"renderMemory(%r, %r) %r\", va, cb, runner)\nself.page().runJavaScript(js, runner)\n+ def __nopcb(self):\n+ pass\n+\ndef contextMenuEvent(self, event):\nif self._canv_curva is not None:\nmenu = vq_ctxmenu.buildContextMenu(self.vw, va=self._canv_curva, parent=self)\n"
}
] |
Python
|
Apache License 2.0
|
vivisect/vivisect
|
viv_loader Cleanup and Polish (#564)
* IMemory bug discovered by Rakuy0
merge fail in ELF parser (regressing certain related sections)
* minor bugfix for VQVivFuncgraphCanvas.renderMemory
|
701,855 |
03.03.2017 11:54:14
| -46,800 |
ed5031cd0b7cbcc6c96fd51d15731d91b330d836
|
Add Gulp front end tooling for scss compilation and live reload
|
[
{
"change_type": "MODIFY",
"old_path": ".gitignore",
"new_path": ".gitignore",
"diff": "@@ -92,3 +92,6 @@ ENV/\n# Rope project settings\n.ropeproject\n+\n+# Installed node modules\n+node_modules/\n"
},
{
"change_type": "MODIFY",
"old_path": "csunplugged/config/settings.py",
"new_path": "csunplugged/config/settings.py",
"diff": "@@ -126,5 +126,5 @@ LOCALE_PATHS = ['locale']\nSTATIC_URL = '/static/'\nSTATICFILES_DIRS = (\n- os.path.join(BASE_DIR, 'static'),\n+ os.path.join(BASE_DIR, 'build'),\n)\n"
},
{
"change_type": "ADD",
"old_path": null,
"new_path": "csunplugged/gulpfile.js",
"diff": "+'use strict';\n+var gulp = require('gulp');\n+var gutil = require('gulp-util');\n+var del = require('del');\n+var uglify = require('gulp-uglify');\n+var gulpif = require('gulp-if');\n+var exec = require('child_process').exec;\n+\n+\n+var notify = require('gulp-notify');\n+\n+var buffer = require('vinyl-buffer');\n+var argv = require('yargs').argv;\n+// sass\n+var sass = require('gulp-sass');\n+var postcss = require('gulp-postcss');\n+var autoprefixer = require('autoprefixer-core');\n+var sourcemaps = require('gulp-sourcemaps');\n+// BrowserSync\n+var browserSync = require('browser-sync');\n+// js\n+var watchify = require('watchify');\n+var browserify = require('browserify');\n+var source = require('vinyl-source-stream');\n+// linting\n+var jshint = require('gulp-jshint');\n+var stylish = require('jshint-stylish');\n+\n+\n+// gulp build --production\n+var production = !!argv.production;\n+// determine if we're doing a build\n+// and if so, bypass the livereload\n+var build = argv._.length ? argv._[0] === 'build' : false;\n+var watch = argv._.length ? argv._[0] === 'watch' : true;\n+\n+// ----------------------------\n+// Error notification methods\n+// ----------------------------\n+var beep = function() {\n+ var os = require('os');\n+ var file = 'gulp/error.wav';\n+ if (os.platform() === 'linux') {\n+ // linux\n+ exec(\"aplay \" + file);\n+ } else {\n+ // mac\n+ console.log(\"afplay \" + file);\n+ exec(\"afplay \" + file);\n+ }\n+};\n+var handleError = function(task) {\n+ return function(err) {\n+ beep();\n+\n+ notify.onError({\n+ message: task + ' failed, check the logs..',\n+ sound: false\n+ })(err);\n+\n+ gutil.log(gutil.colors.bgRed(task + ' error:'), gutil.colors.red(err));\n+ };\n+};\n+// --------------------------\n+// CUSTOM TASK METHODS\n+// --------------------------\n+var tasks = {\n+ // --------------------------\n+ // Delete build folder\n+ // --------------------------\n+ clean: function(cb) {\n+ del(['build/'], cb);\n+ },\n+ // --------------------------\n+ // Copy static images\n+ // --------------------------\n+ images: function() {\n+ return gulp.src('./static/img/**/*')\n+ .pipe(gulp.dest('build/img'));\n+ },\n+ // --------------------------\n+ // CSS\n+ // --------------------------\n+ css: function() {\n+ gulp.src('static/css/**/*.css')\n+ .pipe(gulp.dest('build/css'));\n+ },\n+ // --------------------------\n+ // JS\n+ // --------------------------\n+ js: function() {\n+ gulp.src('static/js/**/*.js')\n+ .pipe(gulp.dest('build/js'));\n+ },\n+ // --------------------------\n+ // SASS (libsass)\n+ // --------------------------\n+ sass: function() {\n+ return gulp.src('./static/scss/*.scss')\n+ // sourcemaps + sass + error handling\n+ .pipe(gulpif(!production, sourcemaps.init()))\n+ .pipe(sass({\n+ sourceComments: !production,\n+ outputStyle: production ? 'compressed' : 'nested'\n+ }))\n+ .on('error', handleError('SASS'))\n+ // generate .maps\n+ .pipe(gulpif(!production, sourcemaps.write({\n+ 'includeContent': false,\n+ 'sourceRoot': '.'\n+ })))\n+ // autoprefixer\n+ .pipe(gulpif(!production, sourcemaps.init({\n+ 'loadMaps': true\n+ })))\n+ .pipe(postcss([autoprefixer({browsers: ['last 2 versions']})]))\n+ // we don't serve the source files\n+ // so include scss content inside the sourcemaps\n+ .pipe(sourcemaps.write({\n+ 'includeContent': true\n+ }))\n+ // write sourcemaps to a specific directory\n+ // give it a file and save\n+ .pipe(gulp.dest('build/css'));\n+ },\n+\n+ // --------------------------\n+ // linting\n+ // --------------------------\n+ lintjs: function() {\n+ return gulp.src([\n+ 'gulpfile.js',\n+ './static/js/index.js',\n+ './static/js/**/*.js'\n+ ]).pipe(jshint())\n+ .pipe(jshint.reporter(stylish))\n+ .on('error', function() {\n+ beep();\n+ });\n+ },\n+};\n+\n+gulp.task('browser-sync', function() {\n+ browserSync.init({\n+ proxy: \"localhost:8000\",\n+ port: process.env.PORT || 3000\n+ });\n+});\n+\n+gulp.task('reload-sass', ['sass'], function(){\n+ browserSync.reload();\n+});\n+gulp.task('reload-js', ['js'], function(){\n+ browserSync.reload();\n+});\n+gulp.task('reload-css', ['css'], function(){\n+ browserSync.reload();\n+});\n+gulp.task('reload-templates', [], function(){\n+ browserSync.reload();\n+});\n+\n+// // --------------------------\n+// // CUSTOMS TASKS\n+// // --------------------------\n+gulp.task('clean', tasks.clean);\n+// // for production we require the clean method on every individual task\n+var req = [];\n+// // individual tasks\n+gulp.task('images', req, tasks.images);\n+gulp.task('js', req, tasks.js);\n+gulp.task('css', req, tasks.css);\n+gulp.task('sass', req, tasks.sass);\n+gulp.task('lint:js', tasks.lintjs);\n+gulp.task('optimize', tasks.optimize);\n+gulp.task('test', tasks.test);\n+\n+// --------------------------\n+// DEV/WATCH TASK\n+// --------------------------\n+gulp.task('watch', ['images', 'css', 'js', 'sass', 'browser-sync'], function() {\n+\n+ // --------------------------\n+ // watch:sass\n+ // --------------------------\n+ gulp.watch('./static/scss/**/*.scss', ['reload-sass']);\n+\n+ // --------------------------\n+ // watch:js\n+ // --------------------------\n+ gulp.watch('./static/js/**/*.js', ['lint:js', 'reload-js']);\n+\n+ // --------------------------\n+ // watch:css\n+ // --------------------------\n+ gulp.watch('./static/css/**/*.css', ['reload-css']);\n+\n+ // --------------------------\n+ // watch:templates\n+ // --------------------------\n+ gulp.watch('./templates/**/*.html', ['reload-templates']);\n+\n+ gutil.log(gutil.colors.bgGreen('Watching for changes...'));\n+});\n+\n+// build task\n+gulp.task('build', [\n+ 'clean',\n+ 'images',\n+ 'css',\n+ 'js',\n+ 'sass'\n+]);\n+\n+gulp.task('default', ['watch']);\n+\n+// gulp (watch) : for development and livereload\n+// gulp build : for a one off development build\n+// gulp build --production : for a minified production build\n"
}
] |
Python
|
MIT License
|
uccser/cs-unplugged
|
Add Gulp front end tooling for scss compilation and live reload
|
701,852 |
20.04.2017 11:59:34
| -43,200 |
8434667df3bda962b4e9e8a37650efd42feb77a7
|
Updated Travis configurations and associated processes.
|
[
{
"change_type": "ADD",
"old_path": null,
"new_path": ".coveragerc",
"diff": "+[run]\n+source = config,dev,general,locale,resources,static,templates,topics,utils\n+omit=*/migrations/*\n+\n+[report]\n+fail_under=50\n"
},
{
"change_type": "MODIFY",
"old_path": ".flake8",
"new_path": ".flake8",
"diff": "@@ -9,6 +9,7 @@ exclude =\n# These files are mostly or completely autogenerated\nmigrations,\n+ manage.py,\ncsunplugged/manage.py,\n# This contains our built project files and documentation\n"
},
{
"change_type": "MODIFY",
"old_path": ".travis.yml",
"new_path": ".travis.yml",
"diff": "@@ -8,10 +8,22 @@ python:\n- \"3.6-dev\" # 3.6 development branch\n# Install dependencies\n-install: pip install flake8\n+install:\n+ - pip install flake8 coverage coveralls\n+ - pip install requirements/base.txt\n+\n+before_script:\n+ cd csunplugged &\n# Runs test suite\n-script: flake8\n+script:\n+ - flake8 --config=../.flake8\n+ - coverage run --rcfile=../.coveragerc manage.py test -v 3\n+ - 'if [ \"$TRAVIS_PULL_REQUEST\" != \"false\" ]; then bash manage.py test --reverse -v 0; fi'\n+\n+after_script:\n+ coverage report --rcfile=../.coveragerc\n+ #coveralls\n# Stop email notifications but post to organisation Slack channel\nnotifications:\n"
},
{
"change_type": "MODIFY",
"old_path": "requirements/base.txt",
"new_path": "requirements/base.txt",
"diff": "@@ -14,6 +14,7 @@ django-environ==0.4.1\nPillow==4.0.0\n# Markdown\n+verto==0.4.0\npython-markdown-math==0.2\n# Models\n"
}
] |
Python
|
MIT License
|
uccser/cs-unplugged
|
Updated Travis configurations and associated processes.
|
701,852 |
20.04.2017 13:19:39
| -43,200 |
a82305411ea33c2a9e063f762185006681992f61
|
Fix reversed tests.
|
[
{
"change_type": "MODIFY",
"old_path": ".travis.yml",
"new_path": ".travis.yml",
"diff": "@@ -18,8 +18,8 @@ before_script:\n# Runs test suite\nscript:\n- flake8 --config=../.flake8\n- - coverage run --rcfile=../.coveragerc manage.py test -v 3\n- - 'if [ \"$TRAVIS_PULL_REQUEST\" != \"false\" ]; then bash manage.py test --reverse -v 0; fi'\n+ - coverage run --rcfile=../.coveragerc manage.py test --pattern \"test_*.py\" -v 3\n+ - 'if [ \"$TRAVIS_PULL_REQUEST\" != \"false\" ]; then manage.py test --pattern \"test_*.py\" --reverse -v 0; fi'\nafter_script:\ncoverage report --rcfile=../.coveragerc\n"
},
{
"change_type": "MODIFY",
"old_path": "csunplugged/tests/topics/models/classroom_resource.py",
"new_path": "csunplugged/tests/topics/models/classroom_resource.py",
"diff": "@@ -11,5 +11,5 @@ class ClassroomResourceModelTest(BaseTestWithDB):\nnew_resource = ClassroomResource.objects.create(\ntext='this is a resource'\n)\n- query_result = ClassroomResource.objects.get(pk=1)\n+ query_result = ClassroomResource.objects.get(text='this is a resource')\nself.assertEqual(query_result, new_resource)\n"
},
{
"change_type": "MODIFY",
"old_path": "csunplugged/tests/topics/models/connected_generated_resource.py",
"new_path": "csunplugged/tests/topics/models/connected_generated_resource.py",
"diff": "@@ -18,5 +18,5 @@ class ConnectedGeneratedResourceModelTest(BaseTestWithDB):\nlesson=lesson,\ndescription='this is a description'\n)\n- query_result = ConnectedGeneratedResource.objects.get(pk=1)\n+ query_result = ConnectedGeneratedResource.objects.get(resource=resource, lesson=lesson)\nself.assertEqual(query_result, new_resource)\n"
},
{
"change_type": "MODIFY",
"old_path": "csunplugged/tests/topics/models/curriculum_area.py",
"new_path": "csunplugged/tests/topics/models/curriculum_area.py",
"diff": "@@ -12,5 +12,6 @@ class CurriculumAreaModelTest(BaseTestWithDB):\nslug='slug',\nname='name',\n)\n- query_result = CurriculumArea.objects.get(pk=1)\n+ print(CurriculumArea.objects.count())\n+ query_result = CurriculumArea.objects.get(slug='slug')\nself.assertEqual(query_result, new_curriculum_area)\n"
}
] |
Python
|
MIT License
|
uccser/cs-unplugged
|
Fix reversed tests.
|
701,852 |
20.04.2017 13:39:42
| -43,200 |
45ceef548793e90c14cfcbaca3a5f4856949e1a5
|
Update travis build for running locally and removed kordac
|
[
{
"change_type": "MODIFY",
"old_path": ".travis.yml",
"new_path": ".travis.yml",
"diff": "@@ -11,6 +11,7 @@ python:\ninstall:\n- pip install flake8 coverage coveralls\n- pip install -r requirements/base.txt\n+ - pip install -r requirements/local.txt\nbefore_script:\ncd csunplugged/\n"
},
{
"change_type": "DELETE",
"old_path": "requirements/kordac.txt",
"new_path": null,
"diff": "-# Required dependencies for Kordac\n-markdown==2.6.8\n-beautifulsoup4==4.5.3\n-Jinja2==2.9.5\n-python-slugify==1.2.1\n"
}
] |
Python
|
MIT License
|
uccser/cs-unplugged
|
Update travis build for running locally and removed kordac
|
701,852 |
20.04.2017 13:43:07
| -43,200 |
d8418663346db1703af0af08b1ee14f917c01a84
|
Include model_mommy in testing requirements
|
[
{
"change_type": "MODIFY",
"old_path": ".travis.yml",
"new_path": ".travis.yml",
"diff": "@@ -12,6 +12,7 @@ install:\n- pip install flake8 coverage coveralls\n- pip install -r requirements/base.txt\n- pip install -r requirements/local.txt\n+ - pip install -r requirements/test.txt\nbefore_script:\ncd csunplugged/\n"
},
{
"change_type": "MODIFY",
"old_path": "requirements/test.txt",
"new_path": "requirements/test.txt",
"diff": "# Check Python style\nflake8==3.3.0\n+\n+# Mock Database\n+model_mommy==1.3.2\n"
}
] |
Python
|
MIT License
|
uccser/cs-unplugged
|
Include model_mommy in testing requirements
|
701,852 |
20.04.2017 13:45:23
| -43,200 |
0a9ffdfba21e657ff9d85f9bec5df5109ac4a67b
|
Cut down dependencies into just the test config.
|
[
{
"change_type": "MODIFY",
"old_path": ".travis.yml",
"new_path": ".travis.yml",
"diff": "@@ -11,7 +11,6 @@ python:\ninstall:\n- pip install flake8 coverage coveralls\n- pip install -r requirements/base.txt\n- - pip install -r requirements/local.txt\n- pip install -r requirements/test.txt\nbefore_script:\n"
},
{
"change_type": "MODIFY",
"old_path": "requirements/test.txt",
"new_path": "requirements/test.txt",
"diff": "# Test dependencies go here.\n-r base.txt\n+# Debugging Tools\n+django-debug-toolbar==1.6\n+django-extensions==1.7.7\n+\n# Check Python style\nflake8==3.3.0\n"
}
] |
Python
|
MIT License
|
uccser/cs-unplugged
|
Cut down dependencies into just the test config.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.