Re: [PATCH 1/2] powerpc/process: fix nested output in show_user_instructions()

[Christophe LEROY]

Le 21/08/2018 Ã 08:27, Michael Ellerman a ÃcritÂ:
> Christophe Leroy <christophe.leroy@xxxxxx> writes:
>
> > When two processes crash at the same time, we sometimes encounter
> > nesting in the middle of a line:
>
> I think "interleaved" is the right word, rather than "nesting".
>
> They're actually (potentially) completely unrelated segfaults, that just
> happen to occur at the same time.
>
> And in fact any output that happens simultaneously will mess things up,
> it doesn't have to be another segfault.

Ok, i reworded in v2.

> > [    4.365317] init[1]: segfault (11) at 0 nip 0 lr 0 code 1
> > [    4.370452] init[1]: code: XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX
> > [    4.372042] init[74]: segfault (11) at 10a74 nip 1000c198 lr 100078c8 code 1 in sh[10000000+14000]
> > [    4.386829] XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX
> > [    4.391542] init[1]: code: XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX XXXXXXXX
> > [    4.400863] init[74]: code: 90010024 bf61000c 91490a7c 3fa01002 3be00000 7d3e4b78 3bbd0c20 3b600000
> > [    4.409867] init[74]: code: 3b9d0040 7c7fe02e 2f830000 419e0028 <89230000> 2f890000 41be001c 4b7f6e79
> >
> > This patch fixes it by preparing complete lines in a buffer and
> > printing it at once.
> >
> > Fixes: 88b0fe1757359 ("powerpc: Add show_user_instructions()")
> > Cc: Murilo Opsfelder Araujo <muriloo@xxxxxxxxxxxxx>
> > Signed-off-by: Christophe Leroy <christophe.leroy@xxxxxx>
> > ---
> >   arch/powerpc/kernel/process.c | 17 +++++++++--------
> >   1 file changed, 9 insertions(+), 8 deletions(-)
> >
> > diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c
> > index 913c5725cdb2..c722ce4ca1c0 100644
> > --- a/arch/powerpc/kernel/process.c
> > +++ b/arch/powerpc/kernel/process.c
> > @@ -1303,32 +1303,33 @@ void show_user_instructions(struct pt_regs *regs)
> >   {
> >   	unsigned long pc;
> >   	int i;
> > +	char buf[96]; /* enough for 8 times 9 + 2 chars */
> > +	int l = 0;
>
> I'm sure your math is right, but still an on-stack buffer with sprintf()
> is a bit scary.
>
> Can you try using seq_buf instead? It is safe against overflow.
>
> eg, something like:
>
> struct seq_buf s;
> char buf[96];
>
> seq_buf_init(&s, buf, sizeof(buf));
> ...
> seq_buf_printf(&s, ...);

Ok, I did that in v2. In the meantime I reworked the loop to avoid this 
uggly test against i % 8 and this duplication of the pr_info() of the 
code line.

Christophe