From dd57023747e33572b31867f890b0d99f55b5cc2f Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Fri, 20 May 2016 04:41:34 +1000 Subject: [PATCH] powerpc: Improve comment explaining why we modify VRSAVE The comment explaining why we modify VRSAVE is misleading, glibc does rely on the behaviour. Update the comment. Signed-off-by: Anton Blanchard Reviewed-by: Cyril Bur Signed-off-by: Michael Ellerman --- arch/powerpc/kernel/vector.S | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/kernel/vector.S b/arch/powerpc/kernel/vector.S index 1c2e7a3..616a6d8 100644 --- a/arch/powerpc/kernel/vector.S +++ b/arch/powerpc/kernel/vector.S @@ -70,10 +70,11 @@ _GLOBAL(load_up_altivec) MTMSRD(r5) /* enable use of AltiVec now */ isync - /* Hack: if we get an altivec unavailable trap with VRSAVE - * set to all zeros, we assume this is a broken application - * that fails to set it properly, and thus we switch it to - * all 1's + /* + * While userspace in general ignores VRSAVE, glibc uses it as a boolean + * to optimise userspace context save/restore. Whenever we take an + * altivec unavailable exception we must set VRSAVE to something non + * zero. Set it to all 1s. See also the programming note in the ISA. */ mfspr r4,SPRN_VRSAVE cmpwi 0,r4,0 -- 2.7.4