On 18/7/22 22:35, kernel test robot wrote:
Greeting,
FYI, we noticed the following commit (built with gcc-11):
commit: f756fe900f17af85c3f4bafc9b9e996bcc0fbeb1 ("[REPOST PATCH v2] vfs: parse: deal with zero length string value")
url: https://github.com/intel-lab-lkp/linux/commits/Ian-Kent/vfs-parse-deal-with-zero-length-string-value/20220708-094030
base: https://git.kernel.org/cgit/linux/kernel/git/viro/vfs.git for-next
patch link: https://lore.kernel.org/linux-fsdevel/165724435867.30814.6980005089665688371.stgit@xxxxxxxxxxxxxxxxx
in testcase: xfstests
version: xfstests-x86_64-c1144bf-1_20220711
with following parameters:
disk: 4HDD
fs: ext2
test: ext4-group-02
ucode: 0xec
test-description: xfstests is a regression test suite for xfs and other files ystems.
test-url: git://git.kernel.org/pub/scm/fs/xfs/xfstests-dev.git
on test machine: 4 threads Intel(R) Core(TM) i5-6500 CPU @ 3.20GHz with 32G memory
caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
If you fix the issue, kindly add following tag
Reported-by: kernel test robot <oliver.sang@xxxxxxxxx>
[ 380.748272][ T5965] EXT4-fs (sda4): mounting ext3 file system using the ext4 subsystem
[ 380.856453][ T5993] EXT4-fs: journaled quota format not specified
[ 380.879248][ T5997] EXT4-fs (sda4): mounting ext3 file system using the ext4 subsystem
[ 380.911204][ T6003] EXT4-fs: journaled quota format not specified
[ 380.924796][ T6007] EXT4-fs: journaled quota format not specified
[ 380.964372][ T6012] general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] SMP KASAN PTI
[ 380.975568][ T6012] KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007]
[ 380.983810][ T6012] CPU: 1 PID: 6012 Comm: mount Tainted: G S I 5.19.0-rc2-00001-gf756fe900f17 #1
[ 380.993786][ T6012] Hardware name: Dell Inc. OptiPlex 7040/0Y7WYT, BIOS 1.1.1 10/07/2015
[ 381.001854][ T6012] RIP: 0010:ext4_parse_param (kbuild/src/consumer/fs/ext4/super.c:2109)
It has to be this:
@@ -2110,12 +2110,12 @@ static int ext4_parse_param(struct fs_context
*fc, struct fs_parameter *param)
switch (token) {
#ifdef CONFIG_QUOTA
case Opt_usrjquota:
- if (!*param->string)
+ if (!param->string || !*param->string)
return unnote_qf_name(fc, USRQUOTA);
else
return note_qf_name(fc, USRQUOTA, param);
case Opt_grpjquota:
- if (!*param->string)
+ if (!param->string || !*param->string)
return unnote_qf_name(fc, GRPQUOTA);
else
return note_qf_name(fc, GRPQUOTA, param);
IMHO it's fragile without the additional check since the file system
has no control over how parameters come to it both in the old and new
systems.
Ian
