Commit Graph

80 Commits

Author SHA1 Message Date
pancake
7e0db0ff8f Fix #8941 - Fix crash in arm.winedbg (null deref) 2017-12-03 13:10:59 +01:00
pancake
fb35122459 Some code cleanup 2017-10-22 23:52:39 +02:00
pancake
94b2fc64d2 Implement RIODescData and adjust RIOMach to work with it
- Handles a magic to ensure the destination struct is the expected one
- Based on r_str_hash, but it must be cached
- Use more r_io_desc_get_pid ()
2017-09-11 01:24:13 +02:00
alvarofe
5d29ca1af6 Fix type matching tests 2017-09-02 22:31:08 +02:00
Fangrui Song
800fed2e31 Fix wt #8394 2017-09-01 19:34:09 +02:00
pancake
949fc79e37 Fix last covs 2017-08-27 13:42:16 +02:00
pancake
16dcd8c10e Fix r_io_seek and coreBlockShift crash 2017-08-25 23:45:37 +02:00
condret
b0c612bf0f kill is_valid_section_offset (#8288) 2017-08-24 00:19:57 +02:00
pancake
9189886f75 Fix infinite loop in afta and other bugs 2017-08-23 00:50:15 +02:00
condret
0161fa0ec2 make read_at bool again and implement RIOAccessLog for accurate error-handling
This reverts commit a9bf4aae2b.
2017-08-22 17:24:04 +00:00
Fangrui Song
a9bf4aae2b Misc fixes to siol (#8262)
* Fix onIterMap (divide-and-conquer like USACO Shaping Regions), check return values of every `op` call
* Change return types of r_io_read_at and friends back to `int`
* Fix some analysis tests and others
* Fix mem leaks
2017-08-22 10:46:48 +02:00
condret
2f201bded0 opsalamance - merge the big siol branch
- Thanks @MaskRay and @condret
2017-08-22 09:42:16 +02:00
alvarofe
2e462ec8f0 Bring RIODesc from SIOL and cleanup 2017-08-18 01:31:57 +00:00
Tzaoh
991a1b8dd9 Replace dh for dL (#7682)
* Replace dh for dL

* Changed places where 'dh' cmd was used.
2017-06-05 14:45:45 +02:00
pancake
e641318973 Fix a bunch of bugs reported by clang-analyzer 2017-04-17 19:59:14 +02:00
pancake
676c7d7176 Some more tiny changes comming from siol into master 2017-03-15 11:06:19 +01:00
Álvaro Felipe Melchor
b245751f33 Arm (#6966)
* Fix issue with arm/thumb bits detection

* r2-indent cmd_print.c
2017-03-09 23:29:48 +01:00
Álvaro Felipe Melchor
05a4ed6607 Added new core binding to seek and set arch and bits
This born to handle arm/thumb since is a pain and we need to
track in which mode we are.
2017-02-07 23:58:35 +01:00
pancake
b0fb7c9261 Fix #6516 - Check null deref before crashing in r_io 2017-01-16 12:15:59 +01:00
Álvaro Felipe Melchor
b8686ed58b Fix issue with thumb and arm
The range was not set up correctly when building the ranges bits
2017-01-08 12:57:09 +01:00
pancake
c355d6ed5e Fix null deref in cbin->info 2017-01-05 10:57:50 +01:00
Álvaro Felipe Melchor
4c1f049f35 fix regression with negative offset in pc 2016-12-28 15:30:45 +01:00
Álvaro Felipe Melchor
87724384d1 added r_cons_break_{push/pop} to handle ^C better
Besides an UAF has been fixed afecting only ELF
2016-11-21 16:56:12 +01:00
pancake
be9df39f6f Fixed 40 null derefs reported by coverity 2016-10-20 15:02:25 +02:00
Álvaro Felipe Melchor
c237950200 Fix #5968 - search string when malloc://X + wx 2016-10-19 23:02:39 +02:00
Sven Steinbauer
2996538700 Fix #5633 - Change x == NULL to correct syntax 2016-09-19 14:44:47 +02:00
pancake
cadb0bd845 Several fixes for gdb:// on avr, still wip 2016-09-13 03:31:54 +02:00
pancake
d01f3832ad Fix #5742 - Honor -n in -d 2016-09-12 23:04:44 +02:00
Chris Rorvick
9b47635f0c remove `next' param from r_core_block_read()
This was originally used to cause a seek to the next block prior to
reading such that successive calls to r_core_block_read() would progress
through memory one block at a time.  This was broken, though, by commit
452669d941 ("more cleanup in r_core_block_read") when when it used
`next' to directly calculate the offset rather than via a seek.

Only one call site remains that attempts to read the next block instead
of the current, and this probably was not even observable due to the
"hacky fix" added in commit 3bfa61946e ("Cleaner pvj, fix tinype load,
and honor 'ao N's").

The current of semantics of `next' appear to be broken and there is very
little dependence on it.  If the original behavior should be restored
anywhere, it would be much better to add a new function, or just do the
seek explicitly, rather than parameterizing r_core_block_read() on it.
2016-08-15 14:26:58 -05:00
Álvaro Felipe Melchor
dc5673ec3a Fix regression with the arm/thumb changes (#5464)
Now we build a list of ranges (RAnalRange) based on anal hints to handle the
bits in a better way in the case of arm/thumb

Now instead of asm.bits that rules the whole binary we can define ranges with
anal hints
2016-08-10 18:49:44 +02:00