* Re: "resource: provide new functions to walk through resources" breaks UML
2014-08-11 18:44 ` Vivek Goyal
@ 2014-08-11 19:18 ` Richard Weinberger
2014-08-11 19:25 ` [uml-devel] " Toralf Förster
2014-08-11 19:48 ` Daniel Walter
2 siblings, 0 replies; 7+ messages in thread
From: Richard Weinberger @ 2014-08-11 19:18 UTC (permalink / raw)
To: Vivek Goyal
Cc: akpm@linux-foundation.org, linux-kernel@vger.kernel.org,
user-mode-linux-devel, sahne
Am 11.08.2014 20:44, schrieb Vivek Goyal:
> On Mon, Aug 11, 2014 at 02:25:56PM +0200, Richard Weinberger wrote:
>> Hi Vivek,
>>
>> Am 11.08.2014 14:22, schrieb Vivek Goyal:
>>> On Sun, Aug 10, 2014 at 06:56:14PM +0200, Richard Weinberger wrote:
>>>> Hi Vivek,
>>>>
>>>> Daniel Walter reported that UML is currently broken.
>>>> The offending commit is:
>>>>
>>>
>>> Hi Richard,
>>>
>>> Thanks for pointing it out. I will look into it.
>>>
>>> Would you have link to the report Daniel sent. I am interested in
>>> looking at backtrace.
>>
>> he reported it on IRC to me.
>>
>> Here is a backtrace:
>> RIP: 0033:[<0000000060039b9f>]
>> RSP: 0000000081459da0 EFLAGS: 00010202
>> RAX: 0000000000000000 RBX: 00000000219b3fff RCX: 000000006010d1d9
>> RDX: 0000000000000001 RSI: 00000000602dfb94 RDI: 0000000081459df8
>> RBP: 0000000081459de0 R08: 00000000601b59f4 R09: ffffffff0000ff00
>> R10: ffffffff0000ff00 R11: 0000000081459e88 R12: 0000000081459df8
>> R13: 00000000219b3fff R14: 00000000602dfb94 R15: 0000000000000000
>> Kernel panic - not syncing: Segfault with no mm
>> CPU: 0 PID: 1 Comm: swapper Not tainted 3.16.0-10454-g58d08e3 #13
>> Stack:
>> 00000000 000080d0 81459df0 219b3fff
>> 81459e70 6010d1d9 ffffffff 6033e010
>> 81459e50 6003a269 81459e30 00000000
>> Call Trace:
>> [<6010d1d9>] ? kclist_add_private+0x0/0xe7
>> [<6003a269>] walk_system_ram_range+0x61/0xb7
>> [<6000e859>] ? proc_kcore_init+0x0/0xf1
>> [<6010d574>] kcore_update_ram+0x4c/0x168
>> [<6010d72e>] ? kclist_add+0x0/0x2e
>> [<6000e943>] proc_kcore_init+0xea/0xf1
>> [<6000e859>] ? proc_kcore_init+0x0/0xf1
>> [<6000e859>] ? proc_kcore_init+0x0/0xf1
>> [<600189f0>] do_one_initcall+0x13c/0x204
>> [<6004ca46>] ? parse_args+0x1df/0x2e0
>> [<6004c82d>] ? parameq+0x0/0x3a
>> [<601b5990>] ? strcpy+0x0/0x18
>> [<60001e1a>] kernel_init_freeable+0x240/0x31e
>> [<6026f1c0>] kernel_init+0x12/0x148
>> [<60019fad>] new_thread_handler+0x81/0xa3
>>
>> RIP is at kernel/resource.c:66.
>
> Hi Richard,
>
> Can you please give the attached patch a try and see if it fixes the UML
> issue.
>
> I noticed that previous code had a for() loop which will not execute
> if p was null. I have resorted back to same code now.
Looks good!
Reported-by: Daniel Walter <sahne@0x90.at>
Tested-by: Richard Weinberger <richard@nod.at>
Thanks,
//richard
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [uml-devel] "resource: provide new functions to walk through resources" breaks UML
2014-08-11 18:44 ` Vivek Goyal
2014-08-11 19:18 ` Richard Weinberger
@ 2014-08-11 19:25 ` Toralf Förster
2014-08-11 19:48 ` Daniel Walter
2 siblings, 0 replies; 7+ messages in thread
From: Toralf Förster @ 2014-08-11 19:25 UTC (permalink / raw)
To: Vivek Goyal, Richard Weinberger
Cc: akpm@linux-foundation.org, linux-kernel@vger.kernel.org,
user-mode-linux-devel, sahne
On 08/11/2014 08:44 PM, Vivek Goyal wrote:
>
> Hi Richard,
>
> Can you please give the attached patch a try and see if it fixes the UML
> issue.
>
> I noticed that previous code had a for() loop which will not execute
> if p was null. I have resorted back to same code now.
>
> Thanks
> Vivek
>
>
> ---
> kernel/resource.c | 11 ++++-------
> 1 file changed, 4 insertions(+), 7 deletions(-)
>
> Index: linux-2.6/kernel/resource.c
> ===================================================================
> --- linux-2.6.orig/kernel/resource.c 2014-08-11 13:36:19.942423192 -0400
> +++ linux-2.6/kernel/resource.c 2014-08-11 13:50:20.529387653 -0400
> @@ -351,15 +351,12 @@ static int find_next_iomem_res(struct re
> end = res->end;
> BUG_ON(start >= end);
>
> - read_lock(&resource_lock);
> -
> - if (first_level_children_only) {
> - p = iomem_resource.child;
> + if (first_level_children_only)
> sibling_only = true;
> - } else
> - p = &iomem_resource;
>
> - while ((p = next_resource(p, sibling_only))) {
> + read_lock(&resource_lock);
> +
> + for (p = iomem_resource.child; p; p = next_resource(p, sibling_only)) {
> if (p->flags != res->flags)
> continue;
> if (name && strcmp(p->name, name))
>
At least at my system (32 bit x86) the patch seems to work (applied on top of v3.16-10473-gc8d6637)
--
Toralf
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: "resource: provide new functions to walk through resources" breaks UML
2014-08-11 18:44 ` Vivek Goyal
2014-08-11 19:18 ` Richard Weinberger
2014-08-11 19:25 ` [uml-devel] " Toralf Förster
@ 2014-08-11 19:48 ` Daniel Walter
2 siblings, 0 replies; 7+ messages in thread
From: Daniel Walter @ 2014-08-11 19:48 UTC (permalink / raw)
To: Vivek Goyal; +Cc: akpm, linux-kernel, user-mode-linux-devel, Richard Weinberger
----- Original Message -----
> From: "Vivek Goyal" <vgoyal@redhat.com>
> To: "Richard Weinberger" <richard@nod.at>
> Cc: akpm@linux-foundation.org, linux-kernel@vger.kernel.org, "user-mode-linux-devel"
> <user-mode-linux-devel@lists.sourceforge.net>, sahne@0x90.at
> Sent: Monday, August 11, 2014 7:44:01 PM
> Subject: Re: "resource: provide new functions to walk through resources" breaks UML
>
> On Mon, Aug 11, 2014 at 02:25:56PM +0200, Richard Weinberger wrote:
> > Hi Vivek,
> >
> > Am 11.08.2014 14:22, schrieb Vivek Goyal:
> > > On Sun, Aug 10, 2014 at 06:56:14PM +0200, Richard Weinberger
> > > wrote:
> > >> Hi Vivek,
> > >>
> > >> Daniel Walter reported that UML is currently broken.
> > >> The offending commit is:
> > >>
> > >
> > > Hi Richard,
> > >
> > > Thanks for pointing it out. I will look into it.
> > >
> > > Would you have link to the report Daniel sent. I am interested in
> > > looking at backtrace.
> >
> > he reported it on IRC to me.
> >
> > Here is a backtrace:
> > RIP: 0033:[<0000000060039b9f>]
> > RSP: 0000000081459da0 EFLAGS: 00010202
> > RAX: 0000000000000000 RBX: 00000000219b3fff RCX: 000000006010d1d9
> > RDX: 0000000000000001 RSI: 00000000602dfb94 RDI: 0000000081459df8
> > RBP: 0000000081459de0 R08: 00000000601b59f4 R09: ffffffff0000ff00
> > R10: ffffffff0000ff00 R11: 0000000081459e88 R12: 0000000081459df8
> > R13: 00000000219b3fff R14: 00000000602dfb94 R15: 0000000000000000
> > Kernel panic - not syncing: Segfault with no mm
> > CPU: 0 PID: 1 Comm: swapper Not tainted 3.16.0-10454-g58d08e3 #13
> > Stack:
> > 00000000 000080d0 81459df0 219b3fff
> > 81459e70 6010d1d9 ffffffff 6033e010
> > 81459e50 6003a269 81459e30 00000000
> > Call Trace:
> > [<6010d1d9>] ? kclist_add_private+0x0/0xe7
> > [<6003a269>] walk_system_ram_range+0x61/0xb7
> > [<6000e859>] ? proc_kcore_init+0x0/0xf1
> > [<6010d574>] kcore_update_ram+0x4c/0x168
> > [<6010d72e>] ? kclist_add+0x0/0x2e
> > [<6000e943>] proc_kcore_init+0xea/0xf1
> > [<6000e859>] ? proc_kcore_init+0x0/0xf1
> > [<6000e859>] ? proc_kcore_init+0x0/0xf1
> > [<600189f0>] do_one_initcall+0x13c/0x204
> > [<6004ca46>] ? parse_args+0x1df/0x2e0
> > [<6004c82d>] ? parameq+0x0/0x3a
> > [<601b5990>] ? strcpy+0x0/0x18
> > [<60001e1a>] kernel_init_freeable+0x240/0x31e
> > [<6026f1c0>] kernel_init+0x12/0x148
> > [<60019fad>] new_thread_handler+0x81/0xa3
> >
> > RIP is at kernel/resource.c:66.
>
> Hi Richard,
>
> Can you please give the attached patch a try and see if it fixes the
> UML
> issue.
>
> I noticed that previous code had a for() loop which will not execute
> if p was null. I have resorted back to same code now.
>
> Thanks
> Vivek
>
>
> ---
> kernel/resource.c | 11 ++++-------
> 1 file changed, 4 insertions(+), 7 deletions(-)
>
> Index: linux-2.6/kernel/resource.c
> ===================================================================
> --- linux-2.6.orig/kernel/resource.c 2014-08-11 13:36:19.942423192
> -0400
> +++ linux-2.6/kernel/resource.c 2014-08-11 13:50:20.529387653 -0400
> @@ -351,15 +351,12 @@ static int find_next_iomem_res(struct re
> end = res->end;
> BUG_ON(start >= end);
>
> - read_lock(&resource_lock);
> -
> - if (first_level_children_only) {
> - p = iomem_resource.child;
> + if (first_level_children_only)
> sibling_only = true;
> - } else
> - p = &iomem_resource;
>
> - while ((p = next_resource(p, sibling_only))) {
> + read_lock(&resource_lock);
> +
> + for (p = iomem_resource.child; p; p = next_resource(p,
> sibling_only)) {
> if (p->flags != res->flags)
> continue;
> if (name && strcmp(p->name, name))
>
Looks good to me, tested it on top of current HEAD (c8d6637d0497d62093dbba0694c7b3a80b79bfe1)
for x86_64.
Thanks,
daniel
^ permalink raw reply [flat|nested] 7+ messages in thread