| Questions | Answers
|------------------------------------------------------|--------------------
| OS/arch/bits (mandatory) | Fedora x86_64
| File format of the file you reverse (mandatory) | PE
| Architecture/bits of the file (mandatory) | x86/32
| r2 -v full output, not truncated (mandatory) | radare2 3.5.0-git 21672 @ linux-x86-64 git.3.4.1-32-g44eb2e26e commit: 44eb2e26e4fef1dd15a877019f417ea0922ea13b build: 2019-04-07__18:59:00
When loading multiple PE binaries via oba, i would like to walk through the symbols using the shell completion.
The problem is that the completion miss some of the symbols
The completion is missing some of the symbols
Look at the asciinema video I recorded.
I want to complete all the symbols which start by sym.nt (ntoskrnl or ntdll), and r2 immediately completes by sym.ntdllxxxx
I can share the binaries if necessary.
Thanks !
maybe this is because there are too many symbols, and the current code in libr/core/core.c uses #define TMP_ARGV_SZ 512 to limit that and avoid hell to happen when the user presses tab with milions of autocompletion options. can you try increasing this? how many flags you have f~?
maybe 4096 is a reasonable number too, but i want to know your specific behaviour in here, or maybe do something smarter and warn when that limit is bypassed
On 7 Apr 2019, at 19:11, Mathieu Tarral notifications@github.com wrote:
Work environment
Questions Answers
OS/arch/bits (mandatory) Fedora x86_64
File format of the file you reverse (mandatory) PE
Architecture/bits of the file (mandatory) x86/32
r2 -v full output, not truncated (mandatory) radare2 3.5.0-git 21672 @ linux-x86-64 git.3.4.1-32-g44eb2e26e commit: 44eb2e2 https://github.com/radare/radare2/commit/44eb2e26e4fef1dd15a877019f417ea0922ea13b build: 2019-04-07__18:59:00
Expected behaviorWhen loading multiple PE binaries via oba, i would like to walk through the symbols using the shell completion.
The problem is that the completion miss some of the symbols
Actual behavior
The completion is missing some of the symbols
Look at the asciinema video I recorded.
I want to complete all the symbols which start by sym.nt (ntoskrnl or ntdll), and r2 immediately completes by sym.ntdllxxxxSteps to reproduce the behavior
asciinema https://asciinema.org/a/YWMwi7hBkkXSNuYFYvolYBeAT
Additional Logs, screenshots, source-code, configuration dump, ...I can share the binaries if necessary.
Thanks !
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub https://github.com/radare/radare2/issues/13651, or mute the thread https://github.com/notifications/unsubscribe-auth/AA3-lgDc3vwbRXrsPKrErLgPVCfivXY6ks5veiargaJpZM4cg4dl.
I have 3750 symbols:
~
[0x005d8c3c]> fs
474 * imports
82 * relocs
13 * resources
21 * sections
0 * segments
410 * strings
3760 * symbols
~
I confirm that manually increasing the limit to 4096 works.
But this should be dynamically allocated, not a statically defined, arbitrary value
can you make a PR adding a config var and making this option change RCore->maxtab instead of being a define?
and we can probably make it 4096 by default
can you make a PR adding a config var and making this option change RCore->maxtab instead of being a define?
I can report issues here and there, but i don't want to get involved in radare2 development at the moment.
I'd rather wait on a radare2 dev to look at this issue.
Thanks
ok will do that when i finish another pr im working on its just 5min
On 8 Apr 2019, at 15:28, Mathieu Tarral notifications@github.com wrote:
can you make a PR adding a config var and making this option change RCore->maxtab instead of being a define?
I can report issues here and there, but i don't want to get involved in radare2 development at the moment.
I'd rather wait on a radare2 dev to look at this issue.Thanks
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub https://github.com/radare/radare2/issues/13651#issuecomment-480831362, or mute the thread https://github.com/notifications/unsubscribe-auth/AA3-ln-FeF6RZwSDZg-bhCJivD62ZAPCks5ve0PvgaJpZM4cg4dl.
should be fine now. ill rename the issue to improve that . thanks!
another thing we can do is eprintf a warning message when the options to autocomplete are larger than the limit, so at least the user knows whats going on