On Wed, 28 Mar 2018, mike.travis@xxxxxxx wrote:
A critical error was found testing the fixed UV4 HUB in that an MMR
address was found to be incorrect. This causes the virtual address
space for accessing the MMIOH1 region to be allocated with the
incorrect size.
Even if this is not intended for stable, despite the diff
--- linux-4.4.orig/arch/x86/include/asm/uv/uv_mmrs.h
+++ linux-4.4/arch/x86/include/asm/uv/uv_mmrs.h
suggesting that this is against a kernel, which receives respiration
treatment, it would be appreciated if you could add a Fixes tag next time.
That spares me the time to figure out which commit introduced the wreckage
and therefore the scope of the fix. Added it myself.
Thanks,
tglx
@@ -3833,7 +3833,7 @@ union uvh_rh_gam_mmioh_overlay_config0_m
#define UV1H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR uv_undefined("UV1H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR")
#define UV2H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR uv_undefined("UV2H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR")
#define UV3H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR 0x1603000UL
-#define UV4H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR 0x483000UL
+#define UV4H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR 0x484000UL
#define UVH_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR ( \
is_uv1_hub() ? UV1H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR : \
is_uv2_hub() ? UV2H_RH_GAM_MMIOH_OVERLAY_CONFIG1_MMR : \
--