[PATCH v3 8/9] powerpc/64: Hard code cache geometry on POWER8
Michael Ellerman
mpe at ellerman.id.au
Fri Feb 3 22:05:58 AEDT 2017
From: Benjamin Herrenschmidt <benh at kernel.crashing.org>
All shipping firmware versions have it wrong in the device-tree
Signed-off-by: Benjamin Herrenschmidt <benh at kernel.crashing.org>
Signed-off-by: Michael Ellerman <mpe at ellerman.id.au>
---
arch/powerpc/kernel/setup_64.c | 20 ++++++++++++++++++--
1 file changed, 18 insertions(+), 2 deletions(-)
v3: Rather than defining ppc_cache_info structs with the p8 values, just call
init_cache_info() directly with the p8 values. This is less code, and means
in the next patch info->assoc will be correctly populated for p8 which it
wasn't in v2 of this patch. (mpe)
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index db18f7b68a1d..364fbffd7e83 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -472,11 +472,27 @@ static bool __init parse_cache_info(struct device_node *np,
void __init initialize_cache_info(void)
{
- struct device_node *cpu, *l2, *l3 = NULL;
+ struct device_node *cpu = NULL, *l2, *l3 = NULL;
+ u32 pvr;
DBG(" -> initialize_cache_info()\n");
- cpu = of_find_node_by_type(NULL, "cpu");
+ /*
+ * All shipping POWER8 machines have a firmware bug that
+ * puts incorrect information in the device-tree. This will
+ * be (hopefully) fixed for future chips but for now hard
+ * code the values if we are running on one of these
+ */
+ pvr = PVR_VER(mfspr(SPRN_PVR));
+ if (pvr == PVR_POWER8 || pvr == PVR_POWER8E ||
+ pvr == PVR_POWER8NVL) {
+ /* size lsize blk sets */
+ init_cache_info(&ppc64_caches.l1i, 0x8000, 128, 128, 32);
+ init_cache_info(&ppc64_caches.l1d, 0x10000, 128, 128, 64);
+ init_cache_info(&ppc64_caches.l2, 0x80000, 128, 0, 512);
+ init_cache_info(&ppc64_caches.l3, 0x800000, 128, 0, 8192);
+ } else
+ cpu = of_find_node_by_type(NULL, "cpu");
/*
* We're assuming *all* of the CPUs have the same
--
2.7.4
More information about the Linuxppc-dev
mailing list