From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: linux-mm@kvack.org
Cc: Christoph Lameter <clameter@engr.sgi.com>,
"Paul E. McKenney" <paulmck@us.ibm.com>,
Nick Piggin <npiggin@suse.de>, Ingo Molnar <mingo@elte.hu>,
Peter Zijlstra <a.p.zijlstra@chello.nl>
Subject: [RFC][PATCH 5/5] x86_64: lockless fault handler
Date: Tue, 06 Mar 2007 02:38:20 +0100 [thread overview]
Message-ID: <20070306014211.523201000@taijtu.programming.kicks-ass.net> (raw)
In-Reply-To: <20070306013815.951032000@taijtu.programming.kicks-ass.net>
[-- Attachment #1: fault-x86_64.patch --]
[-- Type: text/plain, Size: 2014 bytes --]
avoid taking the rw-sem
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
---
arch/x86_64/mm/fault.c | 17 +++++------------
1 file changed, 5 insertions(+), 12 deletions(-)
Index: linux-2.6/arch/x86_64/mm/fault.c
===================================================================
--- linux-2.6.orig/arch/x86_64/mm/fault.c
+++ linux-2.6/arch/x86_64/mm/fault.c
@@ -344,7 +344,6 @@ asmlinkage void __kprobes do_page_fault(
tsk = current;
mm = tsk->mm;
- prefetchw(&mm->mmap_sem);
/* get the address */
__asm__("movq %%cr2,%0":"=r" (address));
@@ -423,14 +422,8 @@ asmlinkage void __kprobes do_page_fault(
* source. If this is invalid we can skip the address space check,
* thus avoiding the deadlock.
*/
- if (!down_read_trylock(&mm->mmap_sem)) {
- if ((error_code & PF_USER) == 0 &&
- !search_exception_tables(regs->rip))
- goto bad_area_nosemaphore;
- down_read(&mm->mmap_sem);
- }
- vma = find_vma(mm, address);
+ vma = find_get_vma(mm, address);
if (!vma)
goto bad_area;
if (likely(vma->vm_start <= address))
@@ -486,7 +479,7 @@ good_area:
goto out_of_memory;
}
- up_read(&mm->mmap_sem);
+ put_vma(vma);
return;
/*
@@ -494,7 +487,7 @@ good_area:
* Fix it, but check if it's kernel or user first..
*/
bad_area:
- up_read(&mm->mmap_sem);
+ put_vma(vma);
bad_area_nosemaphore:
/* User mode accesses just cause a SIGSEGV */
@@ -579,7 +572,7 @@ no_context:
* us unable to handle the page fault gracefully.
*/
out_of_memory:
- up_read(&mm->mmap_sem);
+ put_vma(vma);
if (is_init(current)) {
yield();
goto again;
@@ -590,7 +583,7 @@ out_of_memory:
goto no_context;
do_sigbus:
- up_read(&mm->mmap_sem);
+ put_vma(vma);
/* Kernel mode? Handle exceptions or die */
if (!(error_code & PF_USER))
--
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
prev parent reply other threads:[~2007-03-06 1:38 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-03-06 1:38 [RFC][PATCH 0/5] Lockless vma lookups Peter Zijlstra
2007-03-06 1:38 ` [RFC][PATCH 1/5] RCU friendly B+tree Peter Zijlstra
2007-03-06 1:38 ` [RFC][PATCH 2/5] mm: use B+tree for vmas Peter Zijlstra
2007-03-06 1:38 ` [RFC][PATCH 3/5] mm: RCUify vma lookup Peter Zijlstra
2007-03-06 2:23 ` Nick Piggin
2007-03-06 8:36 ` Nick Piggin
2007-03-08 17:31 ` Ingo Molnar
2007-03-06 12:31 ` Peter Zijlstra
2007-03-06 12:35 ` Peter Zijlstra
2007-03-06 1:38 ` [RFC][PATCH 4/5] i386: lockless fault handler Peter Zijlstra
2007-03-06 1:38 ` Peter Zijlstra [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=20070306014211.523201000@taijtu.programming.kicks-ass.net \
--to=a.p.zijlstra@chello.nl \
--cc=clameter@engr.sgi.com \
--cc=linux-mm@kvack.org \
--cc=mingo@elte.hu \
--cc=npiggin@suse.de \
--cc=paulmck@us.ibm.com \
/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