Re: [sparc64] sigbus in e2fsck
On Tue, Aug 30, 2016 at 11:17:07PM +0300, Anatoly Pugachev wrote:
> On Tue, Aug 30, 2016 at 10:16 PM, Theodore Ts'o <tytso@mit.edu> wrote:
> > On Tue, Aug 30, 2016 at 06:12:39PM +0300, Anatoly Pugachev wrote:
> >>
> >> (gdb) p bh->b_data
> >> $1 = "\300;9\230\000\000\000\005\000\000\253\204\000\000\000\070\000\000\000\000\000\000$\022\000\000\000\000\000\000$<\000\000\000\000\000\000$\270\000\000\000\000\000\000$]\000\000\000\000\000\000$\024",
> >> '\000' <repeats 967 times>
> >> (gdb) p offset
> >> $2 = 16
> >> (gdb) p *bh->b_data
> >> $3 = -64 '\300'
> >> (gdb) p *(bh->b_data+offset)
> >> $6 = 0 '\000'
> >
> > Can you give us "p &bh->b_data" (so we can get the starting address of
> > b_data to make sure it's aligned) and "p offset" (so we can check and
> > make sure offset is sane)?
>
> (gdb) p &bh->b_data
> $7 = (char (*)[1024]) 0x2e9b9c
> (gdb) p offset
> $8 = 16
AFAICT, each bh is malloc'd via e2fsck_allocate_memory and nothing seems
to guarantee that the char b_data[1024] will be aligned to a multiple of
8 (it certainly isn't on x64), so I guess this isn't much of a surprise.
We could change b_data to a pointer and then posix_memalign it.
--D
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
Reply to: