<div dir="ltr"><div class="gmail_default" style><br></div><div class="gmail_default" style>Some paths want to check a spinlock is held, others want to check if its not held, it depends on the context.</div><div class="gmail_default" style>
<br></div><div class="gmail_default" style>So returning 1 on UP would break a bunch of code as well.</div><div class="gmail_extra"><br><br><div class="gmail_quote">On Mon, Jan 14, 2013 at 12:18 AM, Jerry Chu <span dir="ltr"><<a href="mailto:hkchu@google.com" target="_blank">hkchu@google.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div class="gmail_extra"><br><br><div class="gmail_quote"><div class="im">On Sun, Jan 13, 2013 at 7:05 PM, Eric Dumazet <span dir="ltr"><<a href="mailto:edumazet@google.com" target="_blank">edumazet@google.com</a>></span> wrote:<br>
</div><div class="im"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><div dir="ltr"><div>Oh well yes, this doesnt quite work on !SMP.</div>
</div></blockquote><div><br></div></div><div>Strange - how would one assert a spin lock is held, and obviously only for SMP? (I almost think <span style="font-family:arial,sans-serif;font-size:13px">arch_spin_is_locked(lock) should be ((void)(lock), 1) for UP for the purpose of assertion...)</span></div>
<div><br></div><div>Also it looks like there are bunch of other places spin_is_locked() assertion is made in the source tree. (Perhaps they are only configured for MP?)</div><div><br></div><div>Thanks,</div>
<div><br></div><div>Jerry</div><div><div class="h5"><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<div dir="ltr"><div><br></div><div>And this kind of bug is frequent....</div><div>
<br></div><div>See following example :</div><div><br></div><div><div>commit b9980cdcf2524c5fe15d8cbae9c97b3ed6385563</div>
<div>Author: Hugh Dickins <<a href="mailto:hughd@google.com" target="_blank">hughd@google.com</a>></div><div>Date: Wed Feb 8 17:13:40 2012 -0800</div><div><br></div>
<div> mm: fix UP THP spin_is_locked BUGs</div><div> </div><div> Fix CONFIG_TRANSPARENT_HUGEPAGE=y CONFIG_SMP=n CONFIG_DEBUG_VM=y</div><div>
CONFIG_DEBUG_SPINLOCK=n kernel: spin_is_locked() is then always false,</div><div> and so triggers some BUGs in Transparent HugePage codepaths.</div><div> </div><div>
asm-generic/bug.h mentions this problem, and provides a WARN_ON_SMP(x);</div><div> but being too lazy to add VM_BUG_ON_SMP, BUG_ON_SMP, WARN_ON_SMP_ONCE,</div><div> VM_WARN_ON_SMP_ONCE, just test NR_CPUS != 1 in the existing VM_BUG_ONs.</div>
<div> </div><div> Signed-off-by: Hugh Dickins <<a href="mailto:hughd@google.com" target="_blank">hughd@google.com</a>></div><div> Cc: Andrea Arcangeli <<a href="mailto:aarcange@redhat.com" target="_blank">aarcange@redhat.com</a>></div>
<div> Cc: <<a href="mailto:stable@vger.kernel.org" target="_blank">stable@vger.kernel.org</a>></div><div> Signed-off-by: Andrew Morton <<a href="mailto:akpm@linux-foundation.org" target="_blank">akpm@linux-foundation.org</a>></div>
<div> Signed-off-by: Linus Torvalds <<a href="mailto:torvalds@linux-foundation.org" target="_blank">torvalds@linux-foundation.org</a>></div><div><br></div><div>diff --git a/mm/huge_memory.c b/mm/huge_memory.c</div>
<div>index b3ffc21..91d3efb 100644</div><div>--- a/mm/huge_memory.c</div><div>+++ b/mm/huge_memory.c</div><div>@@ -2083,7 +2083,7 @@ static void collect_mm_slot(struct mm_slot *mm_slot)</div>
<div> {</div><div> struct mm_struct *mm = mm_slot->mm;</div><div> </div><div>- VM_BUG_ON(!spin_is_locked(&khugepaged_mm_lock));</div>
<div>+ VM_BUG_ON(NR_CPUS != 1 && !spin_is_locked(&khugepaged_mm_lock));</div><div><br></div></div><div><br></div></div><div class="gmail_extra"><br><br><div class="gmail_quote"><div>
On Sun, Jan 13, 2013 at 1:39 PM, Felix Fietkau <span dir="ltr"><<a href="mailto:nbd@openwrt.org" target="_blank">nbd@openwrt.org</a>></span> wrote:<br></div><div><div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<div>On 2013-01-13 7:03 PM, Eric Dumazet wrote:<br>
> I suspect a bug in the spin_is_locked() implementation on your arch, as<br>
> he socket lock should be held at this point.<br>
</div>I don't think this is an arch implementation bug, this probably happens<br>
on all !SMP systems. See this bit from include/linux/spinlock_up.h:<br>
<br>
#define arch_spin_is_locked(lock) ((void)(lock), 0)<br>
<span><font color="#888888"><br>
- Felix<br>
<br>
</font></span></blockquote></div></div></div><br></div>
</blockquote></div></div></div><br></div></div>
</blockquote></div><br></div></div>