From: Mike Travis <travis@sgi.com>
To: Ingo Molnar <mingo@elte.hu>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Thomas Gleixner <tglx@linutronix.de>,
"H. Peter Anvin" <hpa@zytor.com>,
linux-kernel@vger.kernel.org
Subject: Re: [PATCH 00/11] x86: cleanup early per cpu variables/accesses v5-folded
Date: Mon, 28 Apr 2008 09:27:43 -0700 [thread overview]
Message-ID: <4815FAFF.8010701@sgi.com> (raw)
In-Reply-To: <20080428151339.GA25626@elte.hu>
Thanks, I'll check it out asap. It does look like a pda was accessed too
early. Having "CONFIG_DEBUG_PER_CPU_MAPS" on might help with this.
Mike
Ingo Molnar wrote:
...
> got this early crash:
>
> Allocating PCI resources starting at 50000000 (gap: 40000000:a0000000)
> PERCPU: Allocating 32864 bytes of per cpu data
> PANIC: early exception 0e rip 10:ffffffff80aa9683 error 2 cr2 8
> Pid: 0, comm: swapper Not tainted 2.6.25-sched-devel.git-x86-latest.git #93
>
> Call Trace:
> PANIC: early exception 0e rip 10:ffffffff8020cd69 error 0 cr2 30
>
> http://redhat.com/~mingo/misc/config-Mon_Apr_28_16_20_05_CEST_2008.bad
> http://redhat.com/~mingo/misc/log-Mon_Apr_28_16_20_05_CEST_2008.bad
>
> ... from the dreaded randconfig tester :)
>
> the crash secondary one in dump_trace(), see below. Seems percpu data
> structures are confused?
>
> Ingo
>
> ---------------------->
> ffffffff8020cd20 <dump_trace>:
> ffffffff8020cd20: 55 push %rbp
> ffffffff8020cd21: 48 89 e5 mov %rsp,%rbp
> ffffffff8020cd24: 41 57 push %r15
> ffffffff8020cd26: 41 56 push %r14
> ffffffff8020cd28: 41 55 push %r13
> ffffffff8020cd2a: 41 54 push %r12
> ffffffff8020cd2c: 53 push %rbx
> ffffffff8020cd2d: 48 83 ec 48 sub $0x48,%rsp
> ffffffff8020cd31: e8 5a e8 ff ff callq ffffffff8020b590 <mcount>
> ffffffff8020cd36: 65 8b 04 25 24 00 00 mov %gs:0x24,%eax
> ffffffff8020cd3d: 00
> ffffffff8020cd3e: 89 c0 mov %eax,%eax
> ffffffff8020cd40: 49 89 d4 mov %rdx,%r12
> ffffffff8020cd43: 48 89 4d a8 mov %rcx,-0x58(%rbp)
> ffffffff8020cd47: 48 8d 14 c5 00 00 00 lea 0x0(,%rax,8),%rdx
> ffffffff8020cd4e: 00
> ffffffff8020cd4f: 4c 89 45 a0 mov %r8,-0x60(%rbp)
> ffffffff8020cd53: 4c 89 4d 98 mov %r9,-0x68(%rbp)
> ffffffff8020cd57: 48 85 ff test %rdi,%rdi
> ffffffff8020cd5a: 48 89 55 b0 mov %rdx,-0x50(%rbp)
> ffffffff8020cd5e: 48 8b 15 ab 0b 88 00 mov 0x880bab(%rip),%rdx # ffffffff80a8d910 <_cpu_pda>
> ffffffff8020cd65: 48 8b 04 c2 mov (%rdx,%rax,8),%rax
> ffffffff8020cd69: 48 8b 48 30 mov 0x30(%rax),%rcx
> ffffffff8020cd6d: 0f 84 8a 03 00 00 je ffffffff8020d0fd <dump_trace+0x3dd>
> ffffffff8020cd73: 48 8b 5f 08 mov 0x8(%rdi),%rbx
> ffffffff8020cd77: 4d 85 e4 test %r12,%r12
> ffffffff8020cd7a: 48 89 5d c0 mov %rbx,-0x40(%rbp)
> ffffffff8020cd7e: 0f 84 57 03 00 00 je ffffffff8020d0db <dump_trace+0x3bb>
> ffffffff8020cd84: 48 83 7d a8 00 cmpq $0x0,-0x58(%rbp)
> ffffffff8020cd89: 75 20 jne ffffffff8020cdab <dump_trace+0x8b>
> ffffffff8020cd8b: 65 48 8b 04 25 00 00 mov %gs:0x0,%rax
> ffffffff8020cd92: 00 00
> ffffffff8020cd94: 48 39 c7 cmp %rax,%rdi
> ffffffff8020cd97: 0f 84 6e 03 00 00 je ffffffff8020d10b <dump_trace+0x3eb>
> ffffffff8020cd9d: 48 8b 87 b0 03 00 00 mov 0x3b0(%rdi),%rax
> ffffffff8020cda4: 48 8b 00 mov (%rax),%rax
> ffffffff8020cda7: 48 89 45 a8 mov %rax,-0x58(%rbp)
> ffffffff8020cdab: 49 89 ce mov %rcx,%r14
> ffffffff8020cdae: c7 45 bc 00 00 00 00 movl $0x0,-0x44(%rbp)
> ffffffff8020cdb5: 48 c7 c1 20 24 af 80 mov $0xffffffff80af2420,%rcx
> ffffffff8020cdbc: 48 89 4d 90 mov %rcx,-0x70(%rbp)
> ffffffff8020cdc0: 48 8b 5d b0 mov -0x50(%rbp),%rbx
> ffffffff8020cdc4: 31 ff xor %edi,%edi
> ffffffff8020cdc6: 48 8b 04 13 mov (%rbx,%rdx,1),%rax
> ffffffff8020cdca: 4c 8b 48 08 mov 0x8(%rax),%r9
> ffffffff8020cdce: 66 90 xchg %ax,%ax
> ffffffff8020cdd0: 89 f8 mov %edi,%eax
> ffffffff8020cdd2: 48 8b 5d 90 mov -0x70(%rbp),%rbx
> ffffffff8020cdd6: 89 f9 mov %edi,%ecx
> ffffffff8020cdd8: 48 8d 34 c5 00 00 00 lea 0x0(,%rax,8),%rsi
> ffffffff8020cddf: 00
> ffffffff8020cde0: 4a 8d 04 0e lea (%rsi,%r9,1),%rax
> ffffffff8020cde4: 48 8b 14 18 mov (%rax,%rbx,1),%rdx
> ffffffff8020cde8: 49 39 d4 cmp %rdx,%r12
> ffffffff8020cdeb: 73 19 jae ffffffff8020ce06 <dump_trace+0xe6>
> ffffffff8020cded: 4c 8d 82 00 f0 ff ff lea -0x1000(%rdx),%r8
> ffffffff8020cdf4: 4d 39 c4 cmp %r8,%r12
> ffffffff8020cdf7: 0f 83 0f 02 00 00 jae ffffffff8020d00c <dump_trace+0x2ec>
> ffffffff8020cdfd: 83 ff 03 cmp $0x3,%edi
> ffffffff8020ce00: 0f 84 e9 00 00 00 je ffffffff8020ceef <dump_trace+0x1cf>
> ffffffff8020ce06: 48 ff c7 inc %rdi
prev parent reply other threads:[~2008-04-28 16:27 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-26 0:15 Mike Travis
2008-04-26 0:15 ` [PATCH 01/11] x86: Modify Kconfig to allow up to 4096 cpus Mike Travis
2008-04-27 10:39 ` Pavel Machek
2008-04-28 13:38 ` Ingo Molnar
2008-04-28 13:54 ` Pavel Machek
2008-04-28 15:52 ` Ingo Molnar
2008-04-28 17:13 ` H. Peter Anvin
2008-04-26 0:15 ` [PATCH 02/11] x86: cleanup early per cpu variables/accesses v4 Mike Travis
2008-04-26 0:15 ` [PATCH 03/11] x86: restore pda nodenumber field Mike Travis
2008-04-26 0:15 ` [PATCH 04/11] x86: remove the static 256k node_to_cpumask_map Mike Travis
2008-04-26 0:15 ` [PATCH 05/11] sched: replace MAX_NUMNODES with nr_node_ids in kernel/sched.c Mike Travis
2008-04-26 0:15 ` [PATCH 06/11] cpu: change some globals to statics in drivers/base/cpu.c v2 Mike Travis
2008-04-26 0:15 ` [PATCH 07/11] x86: remove static boot_cpu_pda array Mike Travis
2008-04-26 0:15 ` [PATCH 08/11] x86: Add performance variants of cpumask operators Mike Travis
2008-04-26 0:15 ` [PATCH 09/11] x86: Use performance variant for_each_cpu_mask_nr Mike Travis
2008-04-26 0:15 ` [PATCH 10/11] x86: Use performance variant next_cpu_nr Mike Travis
2008-04-26 0:15 ` [PATCH 11/11] net: Pass reference to cpumask variable in net/sunrpc/svc.c Mike Travis
2008-04-28 13:42 ` [PATCH 00/11] x86: cleanup early per cpu variables/accesses v5-folded Ingo Molnar
2008-04-28 15:13 ` Ingo Molnar
2008-04-28 16:27 ` Mike Travis [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4815FAFF.8010701@sgi.com \
--to=travis@sgi.com \
--cc=akpm@linux-foundation.org \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome