From 7c7f1547b627092737493f0781780af85cc9b1a4 Mon Sep 17 00:00:00 2001 From: Dave Hansen Date: Thu, 7 Aug 2014 10:58:41 -0700 Subject: x86/mm: Fix RCU splat from new TLB tracepoints Dave Jones reported seeing a bug from one of my TLB tracepoints: http://lkml.kernel.org/r/20140806181801.GA4605@redhat.com According to Paul McKenney, the right way to fix this is adding an _rcuidle suffix to the tracepoint. http://lkml.kernel.org/r/20140807065055.GA5821@linux.vnet.ibm.com This patch does just that. Reported-by: Dave Jones , Signed-off-by: Dave Hansen Reviewed-by: Paul E. McKenney Cc: Dave Hansen Cc: Linus Torvalds Link: http://lkml.kernel.org/r/20140807175841.5C92D878@viggo.jf.intel.com Signed-off-by: Ingo Molnar --- arch/x86/mm/tlb.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c index 1fe33987de02..98b7976d44af 100644 --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -49,7 +49,13 @@ void leave_mm(int cpu) if (cpumask_test_cpu(cpu, mm_cpumask(active_mm))) { cpumask_clear_cpu(cpu, mm_cpumask(active_mm)); load_cr3(swapper_pg_dir); - trace_tlb_flush(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL); + /* + * This gets called in the idle path where RCU + * functions differently. Tracing normally + * uses RCU, so we have to call the tracepoint + * specially here. + */ + trace_tlb_flush_rcuidle(TLB_FLUSH_ON_TASK_SWITCH, TLB_FLUSH_ALL); } } EXPORT_SYMBOL_GPL(leave_mm); -- cgit v1.2.3 From 86426851c38d3fe84dee34d7daa71d26c174d409 Mon Sep 17 00:00:00 2001 From: Jeremiah Mahler Date: Sat, 9 Aug 2014 00:38:33 -0700 Subject: x86/mm: Fix sparse 'tlb_single_page_flush_ceiling' warning and make the variable read-mostly A sparse warning is generated about 'tlb_single_page_flush_ceiling' not being declared. arch/x86/mm/tlb.c:177:15: warning: symbol 'tlb_single_page_flush_ceiling' was not declared. Should it be static? Since it isn't used anywhere outside this file, fix the warning by making it static. Also, optimize the use of this variable by adding the __read_mostly directive, as suggested by David Rientjes. Suggested-by: David Rientjes Signed-off-by: Jeremiah Mahler Cc: Dave Hansen Cc: Rik van Riel Cc: Mel Gorman Link: http://lkml.kernel.org/r/1407569913-4035-1-git-send-email-jmmahler@gmail.com Signed-off-by: Ingo Molnar --- arch/x86/mm/tlb.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c index 98b7976d44af..ee61c36d64f8 100644 --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -180,7 +180,7 @@ void flush_tlb_current_task(void) * * This is in units of pages. */ -unsigned long tlb_single_page_flush_ceiling = 33; +static unsigned long tlb_single_page_flush_ceiling __read_mostly = 33; void flush_tlb_mm_range(struct mm_struct *mm, unsigned long start, unsigned long end, unsigned long vmflag) -- cgit v1.2.3 From fb21b84e7f809ef04b1e5aed5d463cf0d4866638 Mon Sep 17 00:00:00 2001 From: Stefan Bader Date: Fri, 15 Aug 2014 10:57:46 +0200 Subject: x86_32, entry: Clean up sysenter_badsys declaration commit 554086d85e "x86_32, entry: Do syscall exit work on badsys (CVE-2014-4508)" introduced a new jump label (sysenter_badsys) but somehow the END statements seem to have gone wrong (at least it feels that way to me). This does not seem to be a fatal problem, but just for the sake of symmetry, change the second syscall_badsys to sysenter_badsys. Signed-off-by: Stefan Bader Link: http://lkml.kernel.org/r/1408093066-31021-1-git-send-email-stefan.bader@canonical.com Acked-by: Andy Lutomirski Signed-off-by: H. Peter Anvin --- arch/x86/kernel/entry_32.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/entry_32.S b/arch/x86/kernel/entry_32.S index 47c410d99f5d..4b0e1dfa2226 100644 --- a/arch/x86/kernel/entry_32.S +++ b/arch/x86/kernel/entry_32.S @@ -683,7 +683,7 @@ END(syscall_badsys) sysenter_badsys: movl $-ENOSYS,%eax jmp sysenter_after_call -END(syscall_badsys) +END(sysenter_badsys) CFI_ENDPROC .macro FIXUP_ESPFIX_STACK -- cgit v1.2.3