Fengguang Wu wrote:c8/2.6.23-rc8-mm2/ Thank you to report it. on_mask() 0 the reason is the WARN_ON(): 390 int smp_call_function_mask(cpumask_t mask, 391 void (*func)(void *), void *info, 392 int wait) 393 { 394 int ret; 395 396 /* Can deadlock when called with interrupts disabled */ 397 WARN_ON(irqs_disabled()); 398 399 spin_lock(&call_lock); 400 ret =3D __smp_call_function_mask(mask, func, info, wait); 401 spin_unlock(&call_lock); 402 return ret; 403 } The patch I sent to Andi didn't include this WARN_ON() and it's why I did= n't find this issue. (see http://lkml.org/lkml/2007/8/24/101) smp_call_function_mask() is called by smp_call_function() which calls a f= unction on all CPU except current. The comment of smp_call_function() specifies: =2E.. * You must not call this function with disabled interrupts or from a * hardware interrupt handler or from a bottom half handler. * Actually there are a few legal cases, like panic. */ So this WARN_ON() is correct, and the caller (global_flush_tlb()) doesn't= follow this rule. I guess this WARN_ON() is only needed when we have current CPU in provide= d mask. So I think we should change: int smp_call_function (void (*func) (void *info), void *info, int nonatom= ic, int wait) { return smp_call_function_mask(cpu_online_map, func, info, wait); } ("cpu_online_map" is a bad choice, comment also specifies: "run a functio= n on all other CPU") to int smp_call_function (void (*func) (void *info), void *info, int nonatom= ic, int wait) { int ret; cpumask_t allbutself; allbutself =3D cpu_online_map; cpu_clear(smp_processor_id(), allbutself); spin_lock(&call_lock); ret =3D __smp_call_function_mask(allbutself, func, info, wait); spin_unlock(&call_lock); return ret; } (which is smp_call_function_mask() without the WARN_ON() and without curr= ent cpu in the mask) Andi, is this correct ? Andrew, should I send a patch implementing this change ? Regards, Laurent --=20 ------------- Laurent.Vivier@bull.net -------------- "Software is hard" - Donald Knuth
| Alexandre Oliva | Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3 |
| Eric W. Biederman | Re: [net-2.6.24][patch 2/2] Dynamically allocate the loopback device |
| Ingo Molnar | Re: containers (was Re: -mm merge plans for 2.6.23) |
| Greg KH | [GIT PATCH] driver core patches against 2.6.24 |
git: | |
| Gerrit Renker | [PATCH 27/37] dccp: Integration of dynamic feature activation - part 2 (server side) |
| David Miller | [GIT]: Networking |
| Michael Riepe | Re: 2.6.27.19 + 28.7: network timeouts for r8169 and 8139too |
