Re: [PATCH 4.9 000/125] 4.9.220-rc1 review

[Guenter Roeck]

On Wed, Apr 22, 2020 at 01:34:30PM -0700, Guenter Roeck wrote:
> On Wed, Apr 22, 2020 at 11:55:17AM +0200, Greg Kroah-Hartman wrote:
> > This is the start of the stable review cycle for the 4.9.220 release.
> > There are 125 patches in this series, all will be posted as a response
> > to this one.  If anyone has any issues with these being applied, please
> > let me know.
> > 
> > Responses should be made by Fri, 24 Apr 2020 09:48:23 +0000.
> > Anything received after that time might be too late.
> > 
> 
> I see a number of unit test crashes in ppc images. Looks like UAF.
> This affects 4.4.y, 4.9.y, and 4.14.y. I'll bisect.
> 

Bisect log attached. I suspect the real culprit is commit a4f91f0de905
("of: unittest: clean up changeset test"), or at least it changes the
code enough for the offending patch not to work in v4.14.y and older.
Either case, reverting upstream commit b3fb36ed694b ("of: unittest:
kmemleak on changeset destroy") fixes the problem and thus needs to
be dropped from v4.4.y, v4.9.y, and v4.14.y.

Guenter

---
# bad: [f6cef26090da1763de1a7fc87205c8442d57bc80] Linux 4.9.220-rc1
# good: [5188957a315f664d46ff58fedecbc0f7503f1b22] Linux 4.9.219
git bisect start 'HEAD' 'v4.9.219'
# good: [0a499a93529e488ccccf83493c48e82b0bdea615] powerpc/fsl_booke: Avoid creating duplicate tlb1 entry
git bisect good 0a499a93529e488ccccf83493c48e82b0bdea615
# bad: [f0b256f314141838039a084b81750d7a9dbc1e16] scsi: ufs: make sure all interrupts are processed
git bisect bad f0b256f314141838039a084b81750d7a9dbc1e16
# good: [9eb52f304182868156a97244618a09898d2dc37b] mac80211_hwsim: Use kstrndup() in place of kasprintf()
git bisect good 9eb52f304182868156a97244618a09898d2dc37b
# good: [cddafffc473da45d9eb419000ab9409824ef8f20] scsi: sg: add sg_remove_request in sg_common_write
git bisect good cddafffc473da45d9eb419000ab9409824ef8f20
# bad: [19976f5df09c19ce1bb0563055586998dda609dd] drm: NULL pointer dereference [null-pointer-deref] (CWE 476) problem
git bisect bad 19976f5df09c19ce1bb0563055586998dda609dd
# bad: [9dbcfb15960da712fc134b5d4a133386721497c5] of: unittest: kmemleak on changeset destroy
git bisect bad 9dbcfb15960da712fc134b5d4a133386721497c5
# good: [52b90d68b8c200b86e66434e0bc86e87510863f3] ALSA: hda: Don't release card at firmware loading error
git bisect good 52b90d68b8c200b86e66434e0bc86e87510863f3
# first bad commit: [9dbcfb15960da712fc134b5d4a133386721497c5] of: unittest: kmemleak on changeset destroy