Re: [LKP] [locking/mutex] e274795ea7 WARNING: CPU: 0 PID: 1 at arch/x86/include/asm/fpu/internal.h:348 __switch_to

On Sat, Feb 18, 2017 at 09:53:07PM +0800, Fengguang Wu wrote: > >>[ 13.377261] Write protecting the kernel text: 15320k > >>[ 13.378910] Write protecting the kernel read-only data: 6316k > >>[ 13.380655] NX-protecting the kernel data: 9256k > >>[ 13.382475] x86/mm: Checked W+X mappings: passed, no W+X pages found. > >>[ 13.384781] ------------[ cut here ]------------ > >>[ 13.386327] WARNING: CPU: 0 PID: 1 at arch/x86/include/asm/fpu/internal.h:348 __switch_to+0x1b6/0x260 > > > >What tree is this? on current next/master that file doesn't have a WARN > >on that line. > > The dmesg comes from the first bad commit e274795ea7 ("locking/mutex: Fix mutex handoff"). > > % git blame -sL 333,350 e274795ea7 arch/x86/include/asm/fpu/internal.h > fd169b05 333) /* > fd169b05 334) * Save processor xstate to xsave area. > fd169b05 335) */ > 8c05f05e 336) static inline void copy_xregs_to_kernel(struct xregs_state *xstate) > fd169b05 337) { > fd169b05 338) u64 mask = -1; > fd169b05 339) u32 lmask = mask; > fd169b05 340) u32 hmask = mask >> 32; > b7106fa0 341) int err; > fd169b05 342) > fd169b05 343) WARN_ON(!alternatives_patched); > fd169b05 344) > b7106fa0 345) XSTATE_XSAVE(xstate, lmask, hmask, err); > fd169b05 346) > 8c05f05e 347) /* We should never fault when copying to a kernel buffer: */ > 8c05f05e 348) WARN_ON_FPU(err); Argh, I'm an idiot.. Please try this one..

--- kernel/fork.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/fork.c b/kernel/fork.c index 68b1706c75d0..a366f636e8af 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -444,7 +444,7 @@ void __init fork_init(void) #ifndef ARCH_MIN_TASKALIGN #define ARCH_MIN_TASKALIGN 0 #endif - int align = min_t(int, L1_CACHE_BYTES, ARCH_MIN_TASKALIGN); + int align = max_t(int, L1_CACHE_BYTES, ARCH_MIN_TASKALIGN); /* create a slab on which task_structs can be allocated */ task_struct_cachep = kmem_cache_create("task_struct",