提交 8c40ad02 编写于 作者: A Andi Kleen 提交者: Andi Kleen

[PATCH] i386: Small cleanup to TLB flush code

- Remove outdated comment
- Use cpu_relax() in a busy loop
Signed-off-by: NAndi Kleen <ak@suse.de>
上级 930f8b8b
...@@ -375,8 +375,7 @@ static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm, ...@@ -375,8 +375,7 @@ static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
/* /*
* i'm not happy about this global shared spinlock in the * i'm not happy about this global shared spinlock in the
* MM hot path, but we'll see how contended it is. * MM hot path, but we'll see how contended it is.
* Temporarily this turns IRQs off, so that lockups are * AK: x86-64 has a faster method that could be ported.
* detected by the NMI watchdog.
*/ */
spin_lock(&tlbstate_lock); spin_lock(&tlbstate_lock);
...@@ -401,7 +400,7 @@ static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm, ...@@ -401,7 +400,7 @@ static void flush_tlb_others(cpumask_t cpumask, struct mm_struct *mm,
while (!cpus_empty(flush_cpumask)) while (!cpus_empty(flush_cpumask))
/* nothing. lockup detection does not belong here */ /* nothing. lockup detection does not belong here */
mb(); cpu_relax();
flush_mm = NULL; flush_mm = NULL;
flush_va = 0; flush_va = 0;
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册