Re: [syzbot] memory leak in __mdiobus_register

[Dongliang Mu]

On Tue, Sep 28, 2021 at 4:15 PM Pavel Skripkin <paskripkin@xxxxxxxxx> wrote:
>
> On 9/28/21 10:42, Dongliang Mu wrote:
> > On Mon, Sep 27, 2021 at 7:44 AM syzbot
> > <syzbot+398e7dc692ddbbb4cfec@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
> >>
> >> Hello,
> >>
> >> syzbot has tested the proposed patch and the reproducer did not trigger any issue:
> >>
> >> Reported-and-tested-by: syzbot+398e7dc692ddbbb4cfec@xxxxxxxxxxxxxxxxxxxxxxxxx
> >>
> >> Tested on:
> >>
> >> commit:         5816b3e6 Linux 5.15-rc3
> >> git tree:       upstream
> >> kernel config:  https://syzkaller.appspot.com/x/.config?x=41799858eb55f380
> >> dashboard link: https://syzkaller.appspot.com/bug?extid=398e7dc692ddbbb4cfec
> >> compiler:       gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian) 2.35.2
> >> patch:          https://syzkaller.appspot.com/x/patch.diff?x=1147b840b00000
> >>
> >> Note: testing is done by a robot and is best-effort only.
> >
> > Hi Pavel,
> >
> > Confirm the patch you posted [1] is the real fix of this bug report.
> >
> > I tested the patch from Yanfei Xu [2] in my local workspace, and the
> > memory leak is still triggered. In addition, I have pushed a patch
> > request for that patch. The result would prove that patch is not
> > working for this bug.
> >
> > BTW, there occur incorrect fix commits on the syzbot dashboard
> > sometimes. Maybe it should be cleaned in the future.
> >
>
>
> Hi, Dongliang,
>
> thank you for confirmation. As I said in reply to [1] Yanfei's patch is
> also correct, but it solves other memory leak in same function.
>

It's fine as I was debugging this case locally.

> AFAIU, if my patch will be applied too there will be 2 fix patches on
> syzkaller bug report page, so no need to remove Yanfei's patch from bug
> report page :)

I don't understand why Dan in other threads said Yanfei's patch is
also working in the bug report. The patch testing request already
shows the same memory leak still triggers. Really confused.

>
>
> > [1] https://lkml.org/lkml/2021/9/27/289
> > [2] https://www.spinics.net/lists/kernel/msg4089781.html
> >
>
>
> With regards,
> Pavel Skripkin