From: Hyunjung Ko <hj351016@gmail.com>
To: Madhavan Srinivasan <maddy@linux.ibm.com>,
Vaibhav Jain <vaibhav@linux.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>,
Nicholas Piggin <npiggin@gmail.com>,
Athira Rajeev <atrajeev@linux.ibm.com>,
Christophe Leroy <chleroy@kernel.org>,
linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org,
Hyunjung Ko <hj351016@gmail.com>,
stable@vger.kernel.org
Subject: [PATCH] powerpc/kvm-hv-pmu: don't allocate under lock_l0_stats in kvmppc_init_hostwide
Date: Wed, 11 Nov 2026 17:45:52 +0900 [thread overview]
Message-ID: <20261111084553.3039506-1-hj351016@gmail.com> (raw)
kvmppc_init_hostwide() builds the hostwide L0-stats guest-state message and
buffer with kvmppc_gsm_new() and kvmppc_gsb_new() while holding
lock_l0_stats, which is taken with spin_lock_irqsave(). Both helpers
allocate with GFP_KERNEL (and kvmppc_gsb_new() does a further GFP_KERNEL
allocation internally), so they can sleep in atomic context, triggering
"BUG: sleeping function called from invalid context" under
CONFIG_DEBUG_ATOMIC_SLEEP.
lock_l0_stats is a spinlock because it also serializes the PMU ->read path
(kvmppc_pmu_event_update()), so it cannot be turned into a mutex. Build the
message and buffer in local variables outside the lock and only take the
lock to publish them, rechecking for a concurrent registration.
Fixes: f0c9c49c5067 ("powerpc/kvm-hv-pmu: Implement GSB message-ops for hostwide counters")
Cc: stable@vger.kernel.org
Signed-off-by: Hyunjung Ko <hj351016@gmail.com>
---
arch/powerpc/perf/kvm-hv-pmu.c | 70 ++++++++++++++++++++++++++----------------
1 file changed, 42 insertions(+), 28 deletions(-)
diff --git a/arch/powerpc/perf/kvm-hv-pmu.c b/arch/powerpc/perf/kvm-hv-pmu.c
index aa72b96b5a8c..deba59cbe234 100644
--- a/arch/powerpc/perf/kvm-hv-pmu.c
+++ b/arch/powerpc/perf/kvm-hv-pmu.c
@@ -314,53 +314,67 @@ static struct kvmppc_gs_msg_ops gsb_ops_l0_stats = {
static int kvmppc_init_hostwide(void)
{
- int rc = 0;
+ struct kvmppc_gs_msg *gsm;
+ struct kvmppc_gs_buff *gsb;
unsigned long flags;
+ int rc;
spin_lock_irqsave(&lock_l0_stats, flags);
-
/* already registered ? */
if (gsm_l0_stats) {
- rc = 0;
- goto out;
+ spin_unlock_irqrestore(&lock_l0_stats, flags);
+ return 0;
}
+ spin_unlock_irqrestore(&lock_l0_stats, flags);
+
+ /*
+ * kvmppc_gsm_new() and kvmppc_gsb_new() allocate with GFP_KERNEL and
+ * may sleep, so build the message and buffer outside lock_l0_stats,
+ * which is a spinlock (it also serializes the PMU ->read path), and
+ * only take the lock to publish them.
+ */
/* setup the Guest state message/buffer to talk to L0 */
- gsm_l0_stats = kvmppc_gsm_new(&gsb_ops_l0_stats, &l0_stats,
- GSM_SEND, GFP_KERNEL);
- if (!gsm_l0_stats) {
- rc = -ENOMEM;
- goto out;
- }
+ gsm = kvmppc_gsm_new(&gsb_ops_l0_stats, &l0_stats, GSM_SEND, GFP_KERNEL);
+ if (!gsm)
+ return -ENOMEM;
/* Populate the Idents */
- kvmppc_gsm_include(gsm_l0_stats, KVMPPC_GSID_L0_GUEST_HEAP);
- kvmppc_gsm_include(gsm_l0_stats, KVMPPC_GSID_L0_GUEST_HEAP_MAX);
- kvmppc_gsm_include(gsm_l0_stats, KVMPPC_GSID_L0_GUEST_PGTABLE_SIZE);
- kvmppc_gsm_include(gsm_l0_stats, KVMPPC_GSID_L0_GUEST_PGTABLE_SIZE_MAX);
- kvmppc_gsm_include(gsm_l0_stats, KVMPPC_GSID_L0_GUEST_PGTABLE_RECLAIM);
+ kvmppc_gsm_include(gsm, KVMPPC_GSID_L0_GUEST_HEAP);
+ kvmppc_gsm_include(gsm, KVMPPC_GSID_L0_GUEST_HEAP_MAX);
+ kvmppc_gsm_include(gsm, KVMPPC_GSID_L0_GUEST_PGTABLE_SIZE);
+ kvmppc_gsm_include(gsm, KVMPPC_GSID_L0_GUEST_PGTABLE_SIZE_MAX);
+ kvmppc_gsm_include(gsm, KVMPPC_GSID_L0_GUEST_PGTABLE_RECLAIM);
/* allocate GSB. Guest/Vcpu Id is ignored */
- gsb_l0_stats = kvmppc_gsb_new(kvmppc_gsm_size(gsm_l0_stats), 0, 0,
- GFP_KERNEL);
- if (!gsb_l0_stats) {
+ gsb = kvmppc_gsb_new(kvmppc_gsm_size(gsm), 0, 0, GFP_KERNEL);
+ if (!gsb) {
rc = -ENOMEM;
- goto out;
+ goto free_gsm;
}
/* ask the ops to fill in the info */
- rc = kvmppc_gsm_fill_info(gsm_l0_stats, gsb_l0_stats);
+ rc = kvmppc_gsm_fill_info(gsm, gsb);
+ if (rc)
+ goto free_gsb;
-out:
- if (rc) {
- if (gsm_l0_stats)
- kvmppc_gsm_free(gsm_l0_stats);
- if (gsb_l0_stats)
- kvmppc_gsb_free(gsb_l0_stats);
- gsm_l0_stats = NULL;
- gsb_l0_stats = NULL;
+ spin_lock_irqsave(&lock_l0_stats, flags);
+ if (gsm_l0_stats) {
+ /* raced with another CPU; use the existing registration */
+ spin_unlock_irqrestore(&lock_l0_stats, flags);
+ rc = 0;
+ goto free_gsb;
}
+ gsm_l0_stats = gsm;
+ gsb_l0_stats = gsb;
spin_unlock_irqrestore(&lock_l0_stats, flags);
+
+ return 0;
+
+free_gsb:
+ kvmppc_gsb_free(gsb);
+free_gsm:
+ kvmppc_gsm_free(gsm);
return rc;
}
--
2.43.0
reply other threads:[~2026-11-11 8:46 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20261111084553.3039506-1-hj351016@gmail.com \
--to=hj351016@gmail.com \
--cc=atrajeev@linux.ibm.com \
--cc=chleroy@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=maddy@linux.ibm.com \
--cc=mpe@ellerman.id.au \
--cc=npiggin@gmail.com \
--cc=stable@vger.kernel.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
all inboxes | Powered by JetHome®