Commit Graph

220 Commits

Author SHA1 Message Date
xorstream 27128a9563 Added mnemonic printing to disasm function. 2015-12-10 13:09:31 +11:00
farmdve 0d98607121 Unmapped memory is not freed.
While uc_mem_unmap does unmap memory regions from Unicorn, it does not
free the memory. It accumulates over time when reusing a single Unicorn
instance.
2015-12-10 00:08:07 +02:00
xorstream eb8d1b58c7 Added MIPS delay slot code hook test.
Tests that the code hook gets called for instructions in the branch delay slot for MIPS cpu.
2015-12-09 18:09:15 +11:00
xorstream 7f04b0f772 Added MIPS delay slot code hook test.
Tests that the code hook gets called for instructions in the branch delay slot for MIPS cpu.
2015-12-09 18:06:53 +11:00
practicalswift 31e32d1734 Crash case: Jump to invalid address (0x0), sparc32 2015-12-02 21:55:13 +01:00
practicalswift 23aa6aa9a6 Add invalid_read_in_tb_flush_x86_64 to .gitignore and Makefile 2015-12-02 07:15:44 +01:00
practicalswift c50bf567c0 Crash case: Invalid read of size 8 in tb_flush_x86_64 2015-12-01 22:41:51 +01:00
practicalswift 7a1067ad4a Crash case: Invalid read of size 4 when tracing (MIPS32). See #282 2015-11-30 23:44:49 +01:00
practicalswift 75a325e8c6 Crash case: Invalid read of size 8 in tb_flush_x86_64. 2015-11-29 23:00:34 +01:00
practicalswift 36948320ce invalid_read_in_cpu_tb_exec_x86_64 → invalid_write_in_cpu_tb_exec_x86_64 2015-11-17 08:44:38 +01:00
practicalswift ad5c2eb9c6 Crash case: Invalid write of size 4 in cpu_tb_exec_x86_64. 2015-11-16 23:19:23 +01:00
Nguyen Anh Quynh edaea7020b x86: on self-modifying code, generate JIT code until end of block. this fixes issue #266 2015-11-16 21:55:42 +08:00
practicalswift 8af4967707 Remove space. 2015-11-15 18:18:04 +01:00
practicalswift 76ed117df6 Cast last two uc_hook_add(…) arguments to uint64_t. 2015-11-15 18:12:10 +01:00
practicalswift 5b34660b3f Add test case to Makefile and .gitignore. Add disassembly. 2015-11-15 17:31:06 +01:00
practicalswift c8fe68fb3d Add test case for issue #266 ("Invalid read of size 4" in cpu_tb_exec). 2015-11-15 00:04:34 +01:00
Willi Ballenthin eb73984763 run_across_bb: include expected vs actual PC in assertion error 2015-11-13 10:14:38 -05:00
Willi Ballenthin d09cc88495 run_across_bb: whitespacing 2015-11-13 09:13:27 -05:00
Willi Ballenthin c97fa4fd8a add test case to run_across_bb
demonstrates that calling emu_start from a BB start to another BB end works fine.
2015-11-13 08:59:36 -05:00
Willi Ballenthin 22d2d3bdbb add test for emulation within and across BB 2015-11-13 00:44:19 -05:00
Nguyen Anh Quynh 9f29fdf442 update regress.sh 2015-11-12 15:30:19 +08:00
Ryan Hileman f93e342177 add mem_map_large regression (for #223) 2015-11-11 22:52:18 -08:00
farmdve c8300587f7 Patch eflags_noset to compile on 32-bit and on 64-bit. 2015-11-10 15:01:23 +02:00
Nguyen Anh Quynh 272293556a do not abort() when memory is insufficient. this fixes issue #244 2015-11-10 11:44:29 +08:00
farmdve 726f45b33d Attempting to set some bits in the Unicorn EFLAGS doesn't work.
When attempting to set all flags except trap flag, the EFLAGS value
should be 0x00244ED7, I've tested this on Windows and here
https://ideone.com/WQAvk1 which is presumably Linux.

Unicorn however has the value 0x00000ED7, bits 11-21 are not set. Bit 21
is the ID bit indicating whether or not CPUID is available.
2015-11-10 00:49:01 +02:00
Nguyen Anh Quynh 4c5ecda908 regress: print with newline for 00opcode_uc_crash.c 2015-11-06 22:24:24 +08:00
Nguyen Anh Quynh de86647dbe regress: fix compiled warning for 00opcode_uc_crash.c 2015-11-05 22:51:10 +08:00
Nguyen Anh Quynh d818e2c485 Merge branch 'master' of https://github.com/unicorn-engine/unicorn 2015-11-05 21:34:55 +08:00
Nguyen Anh Quynh 4692f3445f Merge pull request #248 from farmdve/master
X86 zero-byte opcode causes Unicorn to crash.
2015-11-05 20:49:04 +08:00
Nguyen Anh Quynh b3dfde8a17 Merge branch 'master' of https://github.com/unicorn-engine/unicorn 2015-11-05 20:41:51 +08:00
farmdve 7f3c567ae5 X86 zero-byte opcode causes Unicorn to crash.
The opcode 0x00 translates as `add byte ptr ds:[eax],al`, which leads to
a segfault in Unicorn.

Using a debugger, I believe the crash is located in
`qemu/translate-all.c` in function `page_flush_tb_1`, more specifically
this code `pd[i].first_tb = NULL;`
2015-11-05 14:41:15 +02:00
Nguyen Anh Quynh 95745eff3b regress: fix compiled warnings in eflags_nosync.c 2015-11-05 20:22:07 +08:00
farmdve 894739515e X86 EFLAGS not synced properly.
these commits should fix my previous ones.
2015-11-04 09:38:57 +02:00
Nguyen Anh Quynh 613ebbb20e Merge pull request #244 from emdel/master
SIGABRT issue
2015-11-04 10:07:48 +08:00
mariano 9f7d1812a3 SIGABRT issue 2015-11-03 12:58:20 -08:00
Nguyen Anh Quynh 68a2a0c7ab regress: add shebang for hook_code_stop_emu.py 2015-11-03 22:45:48 +08:00
practicalswift 696c58f9f0 Add test case for segfault-on-emu_stop() bug.
The following code segfaults:
unicorn.Uc(unicorn.UC_ARCH_X86, unicorn.UC_MODE_64).emu_stop()

Tested under Linux and OS X.
2015-11-03 14:11:49 +01:00
Nguyen Anh Quynh f5cec3815a Merge pull request #240 from williballenthin/test-hook-code-stop-emu
add hook_code_stop_emu.py test for issue #232
2015-11-03 09:48:58 +08:00
practicalswift 29903c07f9 Fix typo. 2015-11-02 23:23:02 +01:00
practicalswift b91df4af23 Test case for issue #236 (potential memory leak) 2015-11-02 23:19:38 +01:00
practicalswift a435307543 Test case for issue #237 (OS X crash when creating 2048:th Uc object) 2015-11-02 23:18:30 +01:00
Willi Ballenthin 11dfaf4dec hook_code_stop_emu.py: show PC before asserting 2015-11-02 10:05:00 -05:00
Willi Ballenthin 1a8ca49db1 hook_code_stop_emu.py: formatting 2015-11-02 10:03:31 -05:00
Willi Ballenthin f02d03dde3 add hook_code_stop_emu.py test for issue #232 2015-11-02 09:54:36 -05:00
Nguyen Anh Quynh 3a36e327ab support memory redirection, so the issue #217 is fixed 2015-10-27 14:37:03 +08:00
Nguyen Anh Quynh cea1cf210d tests: mips_kseg0_1.c prints out friendly error message rather than just error code 2015-10-27 12:36:03 +08:00
Nguyen Anh Quynh 7553c9c1c2 tests: add mips_kseg0_1 to Makefile to compile it 2015-10-27 11:25:34 +08:00
Nguyen Anh Quynh 05d89da496 Merge pull request #222 from pkooiman/master
Test for executing from  MIPS kseg0/kseg1 memory ranges
2015-10-27 08:44:32 +08:00
Peter Kooiman 0bad2a8a90 Added test for executing from MIPS kseg0/ksegq1 memory ranges to tests/regress 2015-10-26 18:37:27 +01:00
gaffe 21ce78924c assert that PC was changed in callback-pc test 2015-10-23 11:28:10 -07:00
Nguyen Anh Quynh dfc8658083 regress: fix compilation warning of mem_double_unmap.c 2015-10-23 11:36:09 +08:00
coco dd56621bbb fix for use after free in case of double unmap 2015-10-21 22:25:49 +02:00
Nguyen Anh Quynh b1d41d414b on some hook events, uc_hook_add() should not allow more than 1 handler. this add UC_ERR_HOOK_EXIST error type 2015-10-10 18:01:47 +08:00
Sean Heelan bb4011efe2 Regression test for issue #165 2015-10-06 12:55:09 +07:00
Nguyen Anh Quynh 9410b8fbe7 regress: add shebang for ensure_typedef_consts_generated.py 2015-10-04 11:21:37 +08:00
Sean Heelan fd6b510038 Added a comment explaining what is being tested for 2015-10-03 12:36:17 +07:00
Sean Heelan f62418fc07 Regression test for issue #161 2015-10-03 12:22:41 +07:00
Nguyen Anh Quynh 9e64cba6ec Rename some hook related enums:
- UC_ERR_READ_INVALID -> UC_ERR_READ_UNMAPPED
 - UC_ERR_WRITE_INVALID -> UC_ERR_WRITE_UNMAPPED
 - UC_ERR_FETCH_INVALID -> UC_ERR_FETCH_UNMAPPED
 - UC_MEM_READ_INVALID -> UC_MEM_READ_UNMAPPED
 - UC_MEM_WRITE_INVALID -> UC_MEM_WRITE_UNMAPPED
 - UC_MEM_FETCH_INVALID -> UC_MEM_FETCH_UNMAPPED
 - UC_HOOK_MEM_READ_INVALID -> UC_HOOK_MEM_READ_UNMAPPED
 - UC_HOOK_MEM_WRITE_INVALID -> UC_HOOK_MEM_WRITE_UNMAPPED
 - UC_HOOK_MEM_FETCH_INVALID -> UC_HOOK_MEM_FETCH_UNMAPPED
 - UC_HOOK_MEM_INVALID -> UC_HOOK_MEM_UNMAPPED

This also renames some newly added macros to use _INVALID postfix:

 - UC_HOOK_MEM_READ_ERR -> UC_HOOK_MEM_READ_INVALID
 - UC_HOOK_MEM_WRITE_ERR -> UC_HOOK_MEM_WRITE_INVALID
 - UC_HOOK_MEM_FETCH_ERR -> UC_HOOK_MEM_FETCH_INVALID
 - UC_HOOK_MEM_ERR -> UC_HOOK_MEM_INVALID

Fixed all the bindings Java, Go & Python.
2015-09-30 14:46:55 +08:00
Nguyen Anh Quynh 4a42041a83 handle 'bad ram pointer' case. this fixes issue #159 2015-09-29 18:22:22 +08:00
Nguyen Anh Quynh 976950d3c3 chmod +x tests/regress/bad_ram.py 2015-09-29 15:24:30 +08:00
mothran 0e258b9587 updated a few small errors in the script 2015-09-28 22:50:27 -07:00
mothran eeebcc5a44 added bad_ram regress 2015-09-28 21:41:38 -07:00
Nguyen Anh Quynh 8b4d274c34 regress: convert some mips tests to use unittest 2015-09-28 11:57:24 +08:00
Nguyen Anh Quynh 2b0b4169bc mips: advance PC for SYSCALL instruction. this fixes issue #157 2015-09-28 10:58:43 +08:00
Ryan Hileman 13be3435c9 add regress for #157 2015-09-27 01:08:46 -07:00
Nguyen Anh Quynh 53ce8f217d mips: handle delay slot better for branch instructions. this should fix issue #155 2015-09-27 15:05:40 +08:00
Ryan Hileman 4b42b4be52 add regress for #155 2015-09-26 12:48:25 -07:00
Nguyen Anh Quynh 90eb8f2e72 This commit continues the PR #111
- Allow to register handler separately for invalid memory access
- Add new memory events for hooking:
   - UC_MEM_READ_INVALID, UC_MEM_WRITE_INVALID, UC_MEM_FETCH_INVALID
   - UC_HOOK_MEM_READ_PROT, UC_HOOK_MEM_WRITE_PROT, UC_HOOK_MEM_FETCH_PROT
- Rename UC_ERR_EXEC_PROT to UC_ERR_FETCH_PROT
- Change API uc_hook_add() so event type @type can be combined from hooking types
2015-09-24 14:18:02 +08:00
Nguyen Anh Quynh 14a01b5186 mips: handle delay slot so do not duplicate calling instruction handler. this fixes issue #133 2015-09-22 11:59:53 +08:00
danghvu 3c1d65ea66 Reorganize test directories 2015-09-21 20:47:45 -05:00