[PATCH v5 2/4] x86/resctrl: Enable non-contiguous CBMs in Intel CAT

From: Maciej Wieczor-Retman
Date: Tue Oct 10 2023 - 06:43:23 EST


The setting for non-contiguous 1s support in Intel CAT is
hardcoded to false. On these systems, writing non-contiguous
1s into the schemata file will fail before resctrl passes
the value to the hardware.

In Intel CAT CPUID.0x10.1:ECX[3] and CPUID.0x10.2:ECX[3] stopped
being reserved and now carry information about non-contiguous 1s
value support for L3 and L2 cache respectively. The CAT
capacity bitmask (CBM) supports a non-contiguous 1s value if
the bit is set.

The exception are Haswell systems where non-contiguous 1s value
support needs to stay disabled since they can't make use of CPUID
for Cache allocation.

Originally-by: Fenghua Yu <fenghua.yu@xxxxxxxxx>
Signed-off-by: Maciej Wieczor-Retman <maciej.wieczor-retman@xxxxxxxxx>
Tested-by: Peter Newman <peternewman@xxxxxxxxxx>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@xxxxxxxxxxxxxxx>
Reviewed-by: Peter Newman <peternewman@xxxxxxxxxx>
Reviewed-by: Reinette Chatre <reinette.chatre@xxxxxxxxx>
Reviewed-by: Babu Moger <babu.moger@xxxxxxx>
---
Changelog v5:
- Add Babu's reviewed-by tag.
- Rephrase the patch message to avoid redundancy with the patch content.

Changelog v4:
- Add Ilpo's reviewed-by tag.
- Add Reinette's reviewed-by tag.

Changelog v3:
- Add Peter's tested-by and reviewed-by tags.
- Change patch subject to mention CBMs. (Babu)

Changelog v2:
- Rewrite part of a comment concerning Haswell. (Reinette)

arch/x86/kernel/cpu/resctrl/core.c | 9 ++++++---
arch/x86/kernel/cpu/resctrl/ctrlmondata.c | 10 ++++++----
arch/x86/kernel/cpu/resctrl/internal.h | 9 +++++++++
3 files changed, 21 insertions(+), 7 deletions(-)

diff --git a/arch/x86/kernel/cpu/resctrl/core.c b/arch/x86/kernel/cpu/resctrl/core.c
index c09e4fdded3c..19e0681f0435 100644
--- a/arch/x86/kernel/cpu/resctrl/core.c
+++ b/arch/x86/kernel/cpu/resctrl/core.c
@@ -152,6 +152,7 @@ static inline void cache_alloc_hsw_probe(void)
r->cache.cbm_len = 20;
r->cache.shareable_bits = 0xc0000;
r->cache.min_cbm_bits = 2;
+ r->cache.arch_has_sparse_bitmasks = false;
r->alloc_capable = true;

rdt_alloc_capable = true;
@@ -267,15 +268,18 @@ static void rdt_get_cache_alloc_cfg(int idx, struct rdt_resource *r)
{
struct rdt_hw_resource *hw_res = resctrl_to_arch_res(r);
union cpuid_0x10_1_eax eax;
+ union cpuid_0x10_x_ecx ecx;
union cpuid_0x10_x_edx edx;
- u32 ebx, ecx;
+ u32 ebx;

- cpuid_count(0x00000010, idx, &eax.full, &ebx, &ecx, &edx.full);
+ cpuid_count(0x00000010, idx, &eax.full, &ebx, &ecx.full, &edx.full);
hw_res->num_closid = edx.split.cos_max + 1;
r->cache.cbm_len = eax.split.cbm_len + 1;
r->default_ctrl = BIT_MASK(eax.split.cbm_len + 1) - 1;
r->cache.shareable_bits = ebx & r->default_ctrl;
r->data_width = (r->cache.cbm_len + 3) / 4;
+ if (boot_cpu_data.x86_vendor == X86_VENDOR_INTEL)
+ r->cache.arch_has_sparse_bitmasks = ecx.split.noncont;
r->alloc_capable = true;
}

@@ -872,7 +876,6 @@ static __init void rdt_init_res_defs_intel(void)

if (r->rid == RDT_RESOURCE_L3 ||
r->rid == RDT_RESOURCE_L2) {
- r->cache.arch_has_sparse_bitmasks = false;
r->cache.arch_has_per_cpu_cfg = false;
r->cache.min_cbm_bits = 1;
} else if (r->rid == RDT_RESOURCE_MBA) {
diff --git a/arch/x86/kernel/cpu/resctrl/ctrlmondata.c b/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
index ab45012288bb..beccb0e87ba7 100644
--- a/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
+++ b/arch/x86/kernel/cpu/resctrl/ctrlmondata.c
@@ -87,10 +87,12 @@ int parse_bw(struct rdt_parse_data *data, struct resctrl_schema *s,

/*
* Check whether a cache bit mask is valid.
- * For Intel the SDM says:
- * Please note that all (and only) contiguous '1' combinations
- * are allowed (e.g. FFFFH, 0FF0H, 003CH, etc.).
- * Additionally Haswell requires at least two bits set.
+ * On Intel CPUs, non-contiguous 1s value support is indicated by CPUID:
+ * - CPUID.0x10.1:ECX[3]: L3 non-contiguous 1s value supported if 1
+ * - CPUID.0x10.2:ECX[3]: L2 non-contiguous 1s value supported if 1
+ *
+ * Haswell does not support a non-contiguous 1s value and additionally
+ * requires at least two bits set.
* AMD allows non-contiguous bitmasks.
*/
static bool cbm_validate(char *buf, u32 *data, struct rdt_resource *r)
diff --git a/arch/x86/kernel/cpu/resctrl/internal.h b/arch/x86/kernel/cpu/resctrl/internal.h
index 85ceaf9a31ac..c47ef2f13e8e 100644
--- a/arch/x86/kernel/cpu/resctrl/internal.h
+++ b/arch/x86/kernel/cpu/resctrl/internal.h
@@ -492,6 +492,15 @@ union cpuid_0x10_3_eax {
unsigned int full;
};

+/* CPUID.(EAX=10H, ECX=ResID).ECX */
+union cpuid_0x10_x_ecx {
+ struct {
+ unsigned int reserved:3;
+ unsigned int noncont:1;
+ } split;
+ unsigned int full;
+};
+
/* CPUID.(EAX=10H, ECX=ResID).EDX */
union cpuid_0x10_x_edx {
struct {
--
2.42.0