On 01/16/2017 09:23 PM, David Miller wrote:
From: Khalid Aziz <[email protected]>
Date: Wed, 11 Jan 2017 09:12:53 -0700

+#define SUN4V_MCD_PRECISE                              \
+       ldxa    [%g0] ASI_SCRATCHPAD, %g2;              \
+       ldx     [%g2 + HV_FAULT_D_ADDR_OFFSET], %g4;    \
+       ldx     [%g2 + HV_FAULT_D_CTX_OFFSET], %g5;     \
+       ba,pt   %xcc, etrap;                            \
+       rd      %pc, %g7;                               \
+       ba,pt   %xcc, sun4v_mcd_detect_precise;         \
+       nop;                                            \
+       nop;

Always indent instructions which sit in the delay slot of a non-annulled
branch, call, or jmpl by one extra space.  For example:

        ba,pt   %xcc, 1f
         nop

+sun4v_mcd_detect_precise:
+       or      %l4, %g0, %o1
+       or      %l5, %g0, %o2
+       call    sun4v_mem_corrupt_detect_precise
+       add     %sp, PTREGS_OFF, %o0

Likewise.

Also, just do a normal "mov x, y" instead of the "or x, %g0, y" thing.

Thanks, David. I will fix it in the next version of this patch.

--
Khalid

Reply via email to