tiff: backport upstream fix for CVE-2022-2867, CVE-2022-2868, CVE-2022-2869
Fixes upstream issue #352 (closed) where a buffer overflow was generated
by an uint underflow in tiffcrop.c computeInputPixelOffsets()
calculating (uint32_t)(0 - 1) around line 5210. In the following
tiffcrop tries to read pixels from the image at an offset far beyond the
file-/buffersize.
The main region checks in computeInputPixelOffsets() are now updated to
avoid uint underflow.
This update fixes also upstream issues #350 (closed) and #351 (closed).
Issue 350 is fixed by checking for not allowed zone input cases like -Z
0:0 in getCropOffsets().
Furthermore upstream issue #335 (moved) (closed) and #336 (closed) (closed) are also
fixed.
CVE: CVE-2022-2867, CVE-2022-2868, CVE-2022-2869
Signed-off-by: Ghassane Ben El Aattar ghassaneb.aattar@huawei.com
Merge request reports
Activity
@ghassaneben rebase needed here too
added 263 commits
-
0f360887...4c0160e8 - 262 commits from branch
eclipse/oniro-core:kirkstone
- df4b8a8c - [PATCH] tiff: Fix upstream issue #352 heap-buffer-overflow by correcting uint32_t underflow.
-
0f360887...4c0160e8 - 262 commits from branch
The patch looks good to me, but the commit subject line is very confusing. You are re-using subject and body from the actual patch for libtiff which do not fit into the context of Oniro.
What I would expect is at least a different subject line for the commit. e.g.
tiff: backport upstream fix for CVE-2022-2867, CVE-2022-2868, CVE-2022-2869
That makes clear what the recipe is we are touching an the purpose of the change.
added 1 commit
- ab99d3d3 - [PATCH] tiff: backport upstream fix for CVE-2022-2867, CVE-2022-2868, CVE-2022-2869
- Resolved by Stefan Schmidt
Please drop the [PATCH] prefix. We are not using that in Oniro, its used for a email-based patch workflow.
added 1 commit
- a40d01a8 - tiff: backport upstream fix for CVE-2022-2867, CVE-2022-2868, CVE-2022-2869
enabled automatic add to merge train when the pipeline for 2fa18ed5 succeeds
mentioned in commit 46533ceb