summaryrefslogtreecommitdiff
path: root/arch
diff options
context:
space:
mode:
authorMark Rutland <mark.rutland@arm.com>2017-01-03 18:27:01 +0000
committerCatalin Marinas <catalin.marinas@arm.com>2017-01-04 16:26:18 +0000
commit9d84fb27fa135c99c9fe3de33628774a336a70a8 (patch)
treec4fa86a746481b12cb6a23dfb407fb0638cead59 /arch
parent6ef4fb387d50fa8f3bffdffc868b57e981cdd709 (diff)
arm64: restore get_current() optimisation
Commit c02433dd6de32f04 ("arm64: split thread_info from task stack") inverted the relationship between get_current() and current_thread_info(), with sp_el0 now holding the current task_struct rather than the current thead_info. The new implementation of get_current() prevents the compiler from being able to optimize repeated calls to either, resulting in a noticeable penalty in some microbenchmarks. This patch restores the previous optimisation by implementing get_current() in the same way as our old current_thread_info(), using a non-volatile asm statement. Acked-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Mark Rutland <mark.rutland@arm.com> Reported-by: Davidlohr Bueso <dbueso@suse.de> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Diffstat (limited to 'arch')
-rw-r--r--arch/arm64/include/asm/current.h10
1 files changed, 9 insertions, 1 deletions
diff --git a/arch/arm64/include/asm/current.h b/arch/arm64/include/asm/current.h
index f2bcbe2d9889..86c404171305 100644
--- a/arch/arm64/include/asm/current.h
+++ b/arch/arm64/include/asm/current.h
@@ -9,9 +9,17 @@
struct task_struct;
+/*
+ * We don't use read_sysreg() as we want the compiler to cache the value where
+ * possible.
+ */
static __always_inline struct task_struct *get_current(void)
{
- return (struct task_struct *)read_sysreg(sp_el0);
+ unsigned long sp_el0;
+
+ asm ("mrs %0, sp_el0" : "=r" (sp_el0));
+
+ return (struct task_struct *)sp_el0;
}
#define current get_current()