* current BK boot failure, d_alloc()
@ 2003-03-24 11:50 Jens Axboe
2003-03-24 12:02 ` Jens Axboe
0 siblings, 1 reply; 4+ messages in thread
From: Jens Axboe @ 2003-03-24 11:50 UTC (permalink / raw)
To: Linux Kernel
Hi,
SBF: Simple Boot Flag extension found and enabled.
SBF: Setting boot flags 0x80
Starting balanced_irq
Enabling SEP on CPU 1
Enabling SEP on CPU 0
Program received signal SIGSEGV, Segmentation fault.
0xc016da64 in d_alloc (parent=0x0, name=0xdff8fea8) at
include/asm/string.h:196
196 __asm__ __volatile__(
(gdb) bt
#0 0xc016da64 in d_alloc (parent=0x0, name=0xdff8fea8)
at include/asm/string.h:196
#1 0xc016dd16 in d_alloc_root (root_inode=0xc176ddc4) at
fs/dcache.c:787
#2 0xc014f46b in shmem_fill_super (sb=0xc176ddc4, data=0x1, silent=0)
at mm/shmem.c:1738
#3 0xc015cabf in get_sb_nodev (fs_type=0x1, flags=-1049174588,
data=0x1,
fill_super=0xc014f330 <shmem_fill_super>) at fs/super.c:586
#4 0xc015cc16 in do_kern_mount (fstype=0x1 <Address 0x1 out of bounds>,
flags=1, name=0xc02c3547 "tmpfs", data=0x1) at fs/super.c:639
#5 0xc015ccb7 in kern_mount (type=0x1) at fs/super.c:665
#6 0xc0346620 in init_tmpfs () at mm/shmem.c:1887
#7 0xc0338992 in do_initcalls () at init/main.c:486
#8 0xc01050f5 in init (unused=0x0) at init/main.c:551
(gdb) print *(struct qstr *) name
$1 = {name = 0x0, len = 1, hash = 0, name_str = 0xdff8feb4 "\230µÿß\200µÿß"}
craps out in memcpy() due to name->name == NULL
--
Jens Axboe
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: current BK boot failure, d_alloc()
2003-03-24 11:50 current BK boot failure, d_alloc() Jens Axboe
@ 2003-03-24 12:02 ` Jens Axboe
2003-03-24 17:08 ` Randy.Dunlap
0 siblings, 1 reply; 4+ messages in thread
From: Jens Axboe @ 2003-03-24 12:02 UTC (permalink / raw)
To: Linux Kernel
On Mon, Mar 24 2003, Jens Axboe wrote:
> Hi,
>
> SBF: Simple Boot Flag extension found and enabled.
> SBF: Setting boot flags 0x80
> Starting balanced_irq
> Enabling SEP on CPU 1
> Enabling SEP on CPU 0
>
> Program received signal SIGSEGV, Segmentation fault.
> 0xc016da64 in d_alloc (parent=0x0, name=0xdff8fea8) at
> include/asm/string.h:196
> 196 __asm__ __volatile__(
> (gdb) bt
> #0 0xc016da64 in d_alloc (parent=0x0, name=0xdff8fea8)
> at include/asm/string.h:196
> #1 0xc016dd16 in d_alloc_root (root_inode=0xc176ddc4) at
> fs/dcache.c:787
> #2 0xc014f46b in shmem_fill_super (sb=0xc176ddc4, data=0x1, silent=0)
> at mm/shmem.c:1738
> #3 0xc015cabf in get_sb_nodev (fs_type=0x1, flags=-1049174588,
> data=0x1,
> fill_super=0xc014f330 <shmem_fill_super>) at fs/super.c:586
> #4 0xc015cc16 in do_kern_mount (fstype=0x1 <Address 0x1 out of bounds>,
> flags=1, name=0xc02c3547 "tmpfs", data=0x1) at fs/super.c:639
> #5 0xc015ccb7 in kern_mount (type=0x1) at fs/super.c:665
> #6 0xc0346620 in init_tmpfs () at mm/shmem.c:1887
> #7 0xc0338992 in do_initcalls () at init/main.c:486
> #8 0xc01050f5 in init (unused=0x0) at init/main.c:551
> (gdb) print *(struct qstr *) name
> $1 = {name = 0x0, len = 1, hash = 0, name_str = 0xdff8feb4 "\230µÿß\200µÿß"}
>
> craps out in memcpy() due to name->name == NULL
smells like a compiler problem, with the following patch:
===== fs/dcache.c 1.43 vs edited =====
--- 1.43/fs/dcache.c Sat Mar 22 05:05:21 2003
+++ edited/fs/dcache.c Mon Mar 24 12:58:19 2003
@@ -784,7 +784,8 @@
struct dentry *res = NULL;
if (root_inode) {
- res = d_alloc(NULL, &(const struct qstr) { "/", 1, 0 });
+ struct qstr name = { .name = "/", .len = 1, .hash = 0 };
+ res = d_alloc(NULL, &name);
if (res) {
res->d_sb = root_inode->i_sb;
res->d_parent = res;
--
Jens Axboe
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: current BK boot failure, d_alloc()
2003-03-24 12:02 ` Jens Axboe
@ 2003-03-24 17:08 ` Randy.Dunlap
2003-03-24 17:21 ` Jens Axboe
0 siblings, 1 reply; 4+ messages in thread
From: Randy.Dunlap @ 2003-03-24 17:08 UTC (permalink / raw)
To: axboe; +Cc: linux-kernel
> On Mon, Mar 24 2003, Jens Axboe wrote:
>> Hi,
>>
>>
[snip]
>> craps out in memcpy() due to name->name == NULL
>
> smells like a compiler problem, with the following patch:
>
> ===== fs/dcache.c 1.43 vs edited =====
> --- 1.43/fs/dcache.c Sat Mar 22 05:05:21 2003
> +++ edited/fs/dcache.c Mon Mar 24 12:58:19 2003
> @@ -784,7 +784,8 @@
> struct dentry *res = NULL;
>
> if (root_inode) {
> - res = d_alloc(NULL, &(const struct qstr) { "/", 1, 0 });
> + struct qstr name = { .name = "/", .len = 1, .hash = 0 };
> + res = d_alloc(NULL, &name);
> if (res) {
> res->d_sb = root_inode->i_sb;
> res->d_parent = res;
>
> --
what compiler, please?
~Randy
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: current BK boot failure, d_alloc()
2003-03-24 17:08 ` Randy.Dunlap
@ 2003-03-24 17:21 ` Jens Axboe
0 siblings, 0 replies; 4+ messages in thread
From: Jens Axboe @ 2003-03-24 17:21 UTC (permalink / raw)
To: Randy.Dunlap; +Cc: linux-kernel
On Mon, Mar 24 2003, Randy.Dunlap wrote:
> > On Mon, Mar 24 2003, Jens Axboe wrote:
> >> Hi,
> >>
> >>
> [snip]
> >> craps out in memcpy() due to name->name == NULL
> >
> > smells like a compiler problem, with the following patch:
> >
> > ===== fs/dcache.c 1.43 vs edited =====
> > --- 1.43/fs/dcache.c Sat Mar 22 05:05:21 2003
> > +++ edited/fs/dcache.c Mon Mar 24 12:58:19 2003
> > @@ -784,7 +784,8 @@
> > struct dentry *res = NULL;
> >
> > if (root_inode) {
> > - res = d_alloc(NULL, &(const struct qstr) { "/", 1, 0 });
> > + struct qstr name = { .name = "/", .len = 1, .hash = 0 };
> > + res = d_alloc(NULL, &name);
> > if (res) {
> > res->d_sb = root_inode->i_sb;
> > res->d_parent = res;
> >
> > --
>
> what compiler, please?
gcc 3.3 pre-release, I cannot reproduce it in userspace though. I'll try
and take a look at the generated code.
--
Jens Axboe
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2003-03-24 17:11 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-03-24 11:50 current BK boot failure, d_alloc() Jens Axboe
2003-03-24 12:02 ` Jens Axboe
2003-03-24 17:08 ` Randy.Dunlap
2003-03-24 17:21 ` Jens Axboe
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox