From: Gautam Menghani <gautam@linux.ibm.com>
To: npiggin@gmail.com, danielhb413@gmail.com, harshpb@linux.ibm.com,
pbonzini@redhat.com
Cc: Gautam Menghani <gautam@linux.ibm.com>,
qemu-ppc@nongnu.org, qemu-devel@nongnu.org, kvm@vger.kernel.org,
vaibhav@linux.ibm.com
Subject: [PATCH v2] hw/ppc/spapr_hcall: Return host mitigation characteristics in KVM mode
Date: Thu, 17 Apr 2025 16:49:03 +0530 [thread overview]
Message-ID: <20250417111907.258723-1-gautam@linux.ibm.com> (raw)
Currently, on a P10 KVM guest, the mitigations seen in the output of
"lscpu" command are different from the host. The reason for this
behaviour is that when the KVM guest makes the "h_get_cpu_characteristics"
hcall, QEMU does not consider the data it received from the host via the
KVM_PPC_GET_CPU_CHAR ioctl, and just uses the values present in
spapr->eff.caps[], which in turn just contain the default values set in
spapr_machine_class_init().
Fix this behaviour by making sure that h_get_cpu_characteristics()
returns the data received from the KVM ioctl for a KVM guest.
Perf impact:
With null syscall benchmark[1], ~45% improvement is observed.
1. Vanilla QEMU
$ ./null_syscall
132.19 ns 456.54 cycles
2. With this patch
$ ./null_syscall
91.18 ns 314.57 cycles
[1]: https://ozlabs.org/~anton/junkcode/null_syscall.c
Signed-off-by: Gautam Menghani <gautam@linux.ibm.com>
---
v1 -> v2:
Handle the case where KVM_PPC_GET_CPU_CHAR ioctl fails
hw/ppc/spapr_hcall.c | 6 ++++++
include/hw/ppc/spapr.h | 1 +
target/ppc/kvm.c | 13 ++++++++++---
3 files changed, 17 insertions(+), 3 deletions(-)
diff --git a/hw/ppc/spapr_hcall.c b/hw/ppc/spapr_hcall.c
index 5e1d020e3d..d6db1bdab8 100644
--- a/hw/ppc/spapr_hcall.c
+++ b/hw/ppc/spapr_hcall.c
@@ -1402,6 +1402,12 @@ static target_ulong h_get_cpu_characteristics(PowerPCCPU *cpu,
uint8_t count_cache_flush_assist = spapr_get_cap(spapr,
SPAPR_CAP_CCF_ASSIST);
+ if (kvm_enabled() && spapr->chars.character) {
+ args[0] = spapr->chars.character;
+ args[1] = spapr->chars.behaviour;
+ return H_SUCCESS;
+ }
+
switch (safe_cache) {
case SPAPR_CAP_WORKAROUND:
characteristics |= H_CPU_CHAR_L1D_FLUSH_ORI30;
diff --git a/include/hw/ppc/spapr.h b/include/hw/ppc/spapr.h
index af4aa1cb0f..c41da8cb82 100644
--- a/include/hw/ppc/spapr.h
+++ b/include/hw/ppc/spapr.h
@@ -280,6 +280,7 @@ struct SpaprMachineState {
Error *fwnmi_migration_blocker;
SpaprWatchdog wds[WDT_MAX_WATCHDOGS];
+ struct kvm_ppc_cpu_char chars;
};
#define H_SUCCESS 0
diff --git a/target/ppc/kvm.c b/target/ppc/kvm.c
index 3efc28f18b..fec7bcc347 100644
--- a/target/ppc/kvm.c
+++ b/target/ppc/kvm.c
@@ -2499,7 +2499,8 @@ bool kvmppc_has_cap_xive(void)
static void kvmppc_get_cpu_characteristics(KVMState *s)
{
- struct kvm_ppc_cpu_char c;
+ SpaprMachineState *spapr = SPAPR_MACHINE(qdev_get_machine());
+ struct kvm_ppc_cpu_char c = {0};
int ret;
/* Assume broken */
@@ -2509,18 +2510,24 @@ static void kvmppc_get_cpu_characteristics(KVMState *s)
ret = kvm_vm_check_extension(s, KVM_CAP_PPC_GET_CPU_CHAR);
if (!ret) {
- return;
+ goto err;
}
ret = kvm_vm_ioctl(s, KVM_PPC_GET_CPU_CHAR, &c);
if (ret < 0) {
- return;
+ goto err;
}
+ spapr->chars = c;
cap_ppc_safe_cache = parse_cap_ppc_safe_cache(c);
cap_ppc_safe_bounds_check = parse_cap_ppc_safe_bounds_check(c);
cap_ppc_safe_indirect_branch = parse_cap_ppc_safe_indirect_branch(c);
cap_ppc_count_cache_flush_assist =
parse_cap_ppc_count_cache_flush_assist(c);
+
+ return;
+
+err:
+ memset(&(spapr->chars), 0, sizeof(struct kvm_ppc_cpu_char));
}
int kvmppc_get_cap_safe_cache(void)
--
2.49.0
next reply other threads:[~2025-04-17 11:19 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-17 11:19 Gautam Menghani [this message]
2025-04-30 11:15 ` [PATCH v2] hw/ppc/spapr_hcall: Return host mitigation characteristics in KVM mode Gautam Menghani
2025-06-17 7:16 ` Gautam Menghani
2025-08-19 6:50 ` Gautam Menghani
2025-08-19 6:55 ` Gautam Menghani
2025-09-11 10:18 ` Vaibhav Jain
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20250417111907.258723-1-gautam@linux.ibm.com \
--to=gautam@linux.ibm.com \
--cc=danielhb413@gmail.com \
--cc=harshpb@linux.ibm.com \
--cc=kvm@vger.kernel.org \
--cc=npiggin@gmail.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-ppc@nongnu.org \
--cc=vaibhav@linux.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox