From: Andrew Jones <andrew.jones@linux.dev>
To: kvm@vger.kernel.org, kvm-riscv@lists.infradead.org
Cc: cade.richard@gmail.com, atishp@rivosinc.com, jamestiotio@gmail.com
Subject: [kvm-unit-tests PATCH v2 3/3] riscv: sbi: Add two in hart_mask IPI test
Date: Wed, 6 Nov 2024 12:38:18 +0100 [thread overview]
Message-ID: <20241106113814.42992-8-andrew.jones@linux.dev> (raw)
In-Reply-To: <20241106113814.42992-5-andrew.jones@linux.dev>
We should ensure that when hart_mask has more than one hartid
that both harts get IPIs with a single call of the IPI function.
Signed-off-by: Andrew Jones <andrew.jones@linux.dev>
---
riscv/sbi.c | 39 ++++++++++++++++++++++++++++++++++++++-
1 file changed, 38 insertions(+), 1 deletion(-)
diff --git a/riscv/sbi.c b/riscv/sbi.c
index cdf8d13cc9cf..8ccdf42f902a 100644
--- a/riscv/sbi.c
+++ b/riscv/sbi.c
@@ -364,10 +364,11 @@ static void check_ipi(void)
int nr_cpus_present = cpumask_weight(&cpu_present_mask);
int me = smp_processor_id();
unsigned long max_hartid = 0;
+ unsigned long hartid1, hartid2;
cpumask_t ipi_receivers;
static prng_state ps;
struct sbiret ret;
- int cpu;
+ int cpu, cpu2;
ps = prng_init(0xDEADBEEF);
@@ -398,6 +399,42 @@ static void check_ipi(void)
ipi_hart_check(&ipi_receivers);
report_prefix_pop();
+ report_prefix_push("two in hart_mask");
+
+ if (nr_cpus_present < 3) {
+ report_skip("3 cpus required");
+ goto end_two;
+ }
+
+ cpu = rand_online_cpu(&ps);
+ hartid1 = cpus[cpu].hartid;
+ hartid2 = 0;
+ for_each_present_cpu(cpu2) {
+ if (cpu2 == cpu || cpu2 == me)
+ continue;
+ hartid2 = cpus[cpu2].hartid;
+ if (__builtin_labs(hartid2 - hartid1) < BITS_PER_LONG)
+ break;
+ }
+ if (cpu2 == nr_cpus) {
+ report_skip("hartids are too sparse");
+ goto end_two;
+ }
+
+ cpumask_clear(&ipi_done);
+ cpumask_clear(&ipi_receivers);
+ cpumask_set_cpu(cpu, &ipi_receivers);
+ cpumask_set_cpu(cpu2, &ipi_receivers);
+ on_cpu_async(cpu, ipi_hart_wait, (void *)d);
+ on_cpu_async(cpu2, ipi_hart_wait, (void *)d);
+ ret = sbi_send_ipi((1UL << __builtin_labs(hartid2 - hartid1)) | 1UL, hartid1 < hartid2 ? hartid1 : hartid2);
+ report(ret.error == SBI_SUCCESS, "ipi returned success");
+ while (!cpumask_equal(&ipi_done, &ipi_receivers))
+ cpu_relax();
+ ipi_hart_check(&ipi_receivers);
+end_two:
+ report_prefix_pop();
+
report_prefix_push("broadcast");
cpumask_clear(&ipi_done);
cpumask_copy(&ipi_receivers, &cpu_present_mask);
--
2.47.0
next prev parent reply other threads:[~2024-11-06 11:38 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-11-06 11:38 [kvm-unit-tests PATCH v2 0/3] riscv: sbi: Add IPI tests Andrew Jones
2024-11-06 11:38 ` [kvm-unit-tests PATCH v2 1/3] riscv: Add sbi_send_ipi_broadcast Andrew Jones
2024-11-06 11:38 ` [kvm-unit-tests PATCH v2 2/3] riscv: sbi: Add IPI extension tests Andrew Jones
2024-11-06 11:38 ` Andrew Jones [this message]
2024-11-06 12:22 ` [kvm-unit-tests PATCH v2 3/3] riscv: sbi: Add two in hart_mask IPI test Andrew Jones
2024-11-11 15:07 ` [kvm-unit-tests PATCH v2 0/3] riscv: sbi: Add IPI tests Andrew Jones
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=20241106113814.42992-8-andrew.jones@linux.dev \
--to=andrew.jones@linux.dev \
--cc=atishp@rivosinc.com \
--cc=cade.richard@gmail.com \
--cc=jamestiotio@gmail.com \
--cc=kvm-riscv@lists.infradead.org \
--cc=kvm@vger.kernel.org \
/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