From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-wm1-f49.google.com (mail-wm1-f49.google.com [209.85.128.49]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 31CCD284B25 for ; Fri, 5 Jun 2026 14:06:46 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.128.49 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1780668408; cv=none; b=q/T+kELvPZv+DBO8hX3U1eKC7jbTOeZ449wwwNv86/lJygUF6MGmTP9a8Fph7t42cwL2xiysBQtaA1QtWeT51e0j0IvVMEu3liJNGVyj45HIDur1PAZJah1H4g0sJ+ufNA9ueBYCLikCjOgoDnvj3/4F3tF5TQQbvItRgRyStT0= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1780668408; c=relaxed/simple; bh=1nNQ2oLmYHhRTdoVFpntDJC8qNTkKJT4QU7NykxcDFU=; h=Message-ID:Date:MIME-Version:Subject:To:Cc:References:From: In-Reply-To:Content-Type; b=pT7Zwv4UJWhDAvQlJSMhIa7PGePSz9r50pPglEWzv3G54Zec6SYqRX9Xt47EUWRStVzPj2xwj8albOXmkcvjXpuYwa7epWR1LgjiBsO/dDb8K9dQtQEWUXSvsvJ9zUJDVS73zq2zCED8uPpXPpB8UtvluoHJP2LgBT7xaYluxfY= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com; spf=pass smtp.mailfrom=gmail.com; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b=GqLUoMur; arc=none smtp.client-ip=209.85.128.49 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=gmail.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=gmail.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b="GqLUoMur" Received: by mail-wm1-f49.google.com with SMTP id 5b1f17b1804b1-490b2b037d2so17543545e9.3 for ; Fri, 05 Jun 2026 07:06:46 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20251104; t=1780668405; x=1781273205; darn=vger.kernel.org; h=content-transfer-encoding:in-reply-to:from:content-language :references:cc:to:subject:user-agent:mime-version:date:message-id :from:to:cc:subject:date:message-id:reply-to; bh=dCcP4TVf3aNgZJZJktNdPuXr0mIH8V5tAXXqh5Ye3C8=; b=GqLUoMurrRg9oJLARKaCmwI0JfTdonmCdcB77yYCDInPOdZo1A1brzQE1BFzfHHv1M OqwtAyZVbvA0AlMIR5qAxfWmOsSlotUSr7pIyv85La4tALAEPVJewpHxU9HuUXTc5mwK hVapZ463wbqAu7d0jr3U/hGAShFGFxKkoY+N7o5y9AgiHCOYYclnUrvVWI6Pj8PWW9NV xX6Y43hM700QecG+oSiAE5B3TO1ffC8+JPtLw7JYRBFSNGlpeta1OX55+hgwuNMG9r/a sWkrp5WTcBuO9iUuS0x3H4X6M8RBeE3SjlAqoP85OtGcklSlL9qNqv952SvVmIqd8b5G zwtw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20251104; t=1780668405; x=1781273205; h=content-transfer-encoding:in-reply-to:from:content-language :references:cc:to:subject:user-agent:mime-version:date:message-id :x-gm-gg:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=dCcP4TVf3aNgZJZJktNdPuXr0mIH8V5tAXXqh5Ye3C8=; b=Fc8C0QnUBKJ9rlelxYrBBljE6TyPOLdvN8sZrqJqFta+kMK2fYmOM4qvrVFE+cwUhA ZHjYD/JUum5/JMFvKpVqnZov2M/b8zm0w6/znsuuyCJw7JJgvbPZ9KOk18W6oJTfYSNd Wfqukr4/T12iA0ftgH4xdBOpAWkx9kI26GkWRszViGiuJclnxYSvMYSFBbRRLDWTJkaG e2pPjaqaz3j4tKGsx+MW5t3ckBmGz3W6Lqudc2eTxhY5xgMTmkUYYpuJUsjoFDLOmCTh xNSBW2bPhGDnjR4lroOzaxbOMwQyrjHlqyycLAr2h9Xweb3MWvQYcEDcFGKwUmUmQ7so LT6Q== X-Forwarded-Encrypted: i=1; AFNElJ/Gf9SzAY8AJwRmKPA7d8OadNm6HjkfH/2Rn6oBmE93w0fr/xK6JoYoj5ebLfjZZMavjGV6g3/0KjUBtRE=@vger.kernel.org X-Gm-Message-State: AOJu0Ywg59weXzvwTW+CCy6eWqbCf/CCQVDNc2I3oancKN+Hj2ECVFhk m8JKQYaZQ5sbfrYDxrqOD+IFdMzDdqeGaQ+P6bUdULfZfexdVzBYeC7j X-Gm-Gg: Acq92OEl4DwBmdvv19BGh/rSm1GgN4SJfsvfby5U6QmJ4ivpMzESDP/P/C2XS7JRNvJ hd6UcIypqWnMFCLfyzdBu60R8lKuVBD41luGictGXqT6m3T2nipF1TJg/sAWHYm7hpddM8kPj+b 5OE7z2htiAy31bYrT1E7XFttvcOgVgEEvHX3wXXxbN4abVilpvTLyzfReSVv7DzQ2tqKpCg/cPf 5fEa/rHrjRLNfWGsKaQl7KxAkQp91RFxMHDaNevVRyhGcWxGHI5Vi6Pgox87lErj5RzJGMs0gmZ u9Xbpw99ZNy6wz8zY7XBQJqfJHOZZyUiKK8VhRfUuEHwIGqNmaacLho+X+bzAKIuHOqxyq49cFb WoJSZKcDTpBIXibAIaY7a8H73Uv5NFjuxNzzlzErPSOljML4pV19a9jIAW51t/zcFYW4i+cxO3K kmItgdxCUbJVtJNjTcCceO3T4LDlLUrZQQNqsrCm9ez95tDmoFhmMhOuFDaZpNVH9/Y1RyjCkPm +KS77fVsXxLThvHubXtRQ== X-Received: by 2002:a05:600c:37c9:b0:490:adb6:7957 with SMTP id 5b1f17b1804b1-490c262dcb5mr59369655e9.33.1780668405124; Fri, 05 Jun 2026 07:06:45 -0700 (PDT) Received: from ?IPV6:2a02:8109:a307:d900:9c30:fc5f:5bdd:2e3f? ([2a02:8109:a307:d900:9c30:fc5f:5bdd:2e3f]) by smtp.gmail.com with ESMTPSA id ffacd0b85a97d-4601f351ac0sm43474029f8f.27.2026.06.05.07.06.44 (version=TLS1_3 cipher=TLS_AES_128_GCM_SHA256 bits=128/128); Fri, 05 Jun 2026 07:06:44 -0700 (PDT) Message-ID: <5ac74b9b-3738-49c7-bf44-70b72a1ec1b1@gmail.com> Date: Fri, 5 Jun 2026 15:06:43 +0100 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH bpf-next v2] selftests/bpf: Run test_perf_branches more than 1 time To: Maya Matuszczyk , Alexei Starovoitov , Daniel Borkmann , Andrii Nakryiko , Martin KaFai Lau , Eduard Zingerman Cc: Matt Bobrowski , bpf@vger.kernel.org, linux-kernel@vger.kernel.org References: <20260605130601.2699172-1-mayamatuszczyk@google.com> Content-Language: en-US From: Mykyta Yatsenko In-Reply-To: <20260605130601.2699172-1-mayamatuszczyk@google.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit On 6/5/26 2:06 PM, Maya Matuszczyk wrote: > test_perf_branches bpf program was only being ran once, and > perf_branches test would fail if the only sample that was passed to > test_perf_branches didn't have branch data. > Fix those failures by running it a few more times and waiting until a valid > sample shows up, and leave the perf event only enabled for the duration of > test spins. > > Signed-off-by: Maya Matuszczyk > --- Change looks good. minor nit: Maybe it makes sense to migrate all CHECKs in this test to ASSERT, though. Acked-by: Mykyta Yatsenko > Changes from v1: > - Switched from CHECK to ASSERT_OK (Alexei Starovoitov) > - bpf -> bpf-next > > v1: https://lore.kernel.org/bpf/20260408095710.4090495-1-mayamatuszczyk@google.com/ > --- > .../selftests/bpf/prog_tests/perf_branches.c | 57 +++++++++++++------ > .../selftests/bpf/progs/test_perf_branches.c | 11 ++-- > 2 files changed, 43 insertions(+), 25 deletions(-) > > diff --git a/tools/testing/selftests/bpf/prog_tests/perf_branches.c b/tools/testing/selftests/bpf/prog_tests/perf_branches.c > index 0a7ef770c487..153afcdc082e 100644 > --- a/tools/testing/selftests/bpf/prog_tests/perf_branches.c > +++ b/tools/testing/selftests/bpf/prog_tests/perf_branches.c > @@ -15,10 +15,6 @@ static void check_good_sample(struct test_perf_branches *skel) > int pbe_size = sizeof(struct perf_branch_entry); > int duration = 0; > > - if (CHECK(!skel->bss->run_cnt, "invalid run_cnt", > - "checked sample validity before prog run")) > - return; > - > if (CHECK(!skel->bss->valid, "output not valid", > "no valid sample from prog")) > return; > @@ -49,10 +45,6 @@ static void check_bad_sample(struct test_perf_branches *skel) > int written_stack = skel->bss->written_stack_out; > int duration = 0; > > - if (CHECK(!skel->bss->run_cnt, "invalid run_cnt", > - "checked sample validity before prog run")) > - return; > - > if (CHECK(!skel->bss->valid, "output not valid", > "no valid sample from prog")) > return; > @@ -73,7 +65,6 @@ static void test_perf_branches_common(int perf_fd, > bool detached = false; > struct bpf_link *link; > volatile int j = 0; > - cpu_set_t cpu_set; > > skel = test_perf_branches__open_and_load(); > if (CHECK(!skel, "test_perf_branches_load", > @@ -85,19 +76,35 @@ static void test_perf_branches_common(int perf_fd, > if (!ASSERT_OK_PTR(link, "attach_perf_event")) > goto out_destroy_skel; > > - /* generate some branches on cpu 0 */ > - CPU_ZERO(&cpu_set); > - CPU_SET(0, &cpu_set); > - err = pthread_setaffinity_np(pthread_self(), sizeof(cpu_set), &cpu_set); > - if (CHECK(err, "set_affinity", "cpu #0, err %d\n", err)) > + err = ioctl(perf_fd, PERF_EVENT_IOC_RESET, 0); > + if (!ASSERT_OK(err, "ioctl(PERF_EVENT_IOC_RESET)")) > goto out_destroy; > > - /* Spin the loop for a while by using a high iteration count, and by > - * checking whether the specific run count marker has been explicitly > - * incremented at least once by the backing perf_event BPF program. > + err = ioctl(perf_fd, PERF_EVENT_IOC_REFRESH, 10); > + if (!ASSERT_OK(err, "ioctl(PERF_EVENT_IOC_REFRESH)")) > + goto out_destroy; > + > + /* Make up a bunch of branch events until we time out or bpf prog sets a flag > + * that it caught a valid event. To be completely sure that required_size_out, > + * written_stack_out or written_global_out writes and reads won't get > + * reordered by anyone in a way that would fail this test, use atomic > + * operations > */ > - for (i = 0; i < 100000000 && !*(volatile int *)&skel->bss->run_cnt; ++i) > + for (i = 0; i < 100000000; ++i) { > + /* This technically should be an atomic load with acquire > + * ordering, but we don't want to clog up memory bus while > + * spinning > + */ > + if (READ_ONCE(skel->bss->valid) > 0) > + break; > ++j; > + } > + > + __atomic_thread_fence(__ATOMIC_ACQUIRE); > + > + err = ioctl(perf_fd, PERF_EVENT_IOC_DISABLE, 0); > + if (!ASSERT_OK(err, "ioctl(PERF_EVENT_IOC_DISABLE)")) > + goto out_destroy; > > test_perf_branches__detach(skel); > detached = true; > @@ -121,6 +128,7 @@ static void test_perf_branches_hw(void) > attr.size = sizeof(attr); > attr.type = PERF_TYPE_HARDWARE; > attr.config = PERF_COUNT_HW_CPU_CYCLES; > + attr.disabled = 1; > attr.freq = 1; > attr.sample_freq = 1000; > attr.sample_type = PERF_SAMPLE_BRANCH_STACK; > @@ -162,6 +170,7 @@ static void test_perf_branches_no_hw(void) > attr.size = sizeof(attr); > attr.type = PERF_TYPE_SOFTWARE; > attr.config = PERF_COUNT_SW_CPU_CLOCK; > + attr.disabled = 1; > attr.freq = 1; > attr.sample_freq = 1000; > pfd = syscall(__NR_perf_event_open, &attr, -1, 0, -1, PERF_FLAG_FD_CLOEXEC); > @@ -175,6 +184,18 @@ static void test_perf_branches_no_hw(void) > > void test_perf_branches(void) > { > + int err, duration = 0; > + cpu_set_t cpu_set; > + > + /* Pin ourselves to cpu0 so that all branches we generate would be > + * also on cpu0 > + */ > + CPU_ZERO(&cpu_set); > + CPU_SET(0, &cpu_set); > + err = pthread_setaffinity_np(pthread_self(), sizeof(cpu_set), &cpu_set); > + if (!ASSERT_OK(err, "pthread_set_affinity")) > + return; > + > if (test__start_subtest("perf_branches_hw")) > test_perf_branches_hw(); > if (test__start_subtest("perf_branches_no_hw")) > diff --git a/tools/testing/selftests/bpf/progs/test_perf_branches.c b/tools/testing/selftests/bpf/progs/test_perf_branches.c > index 05ac9410cd68..9806414cebc4 100644 > --- a/tools/testing/selftests/bpf/progs/test_perf_branches.c > +++ b/tools/testing/selftests/bpf/progs/test_perf_branches.c > @@ -8,7 +8,6 @@ > #include > > int valid = 0; > -int run_cnt = 0; > int required_size_out = 0; > int written_stack_out = 0; > int written_global_out = 0; > @@ -25,13 +24,11 @@ int perf_branches(void *ctx) > __u64 entries[4 * 3] = {0}; > int required_size, written_stack, written_global; > > - ++run_cnt; > - > /* write to stack */ > written_stack = bpf_read_branch_records(ctx, entries, sizeof(entries), 0); > /* ignore spurious events */ > if (!written_stack) > - return 1; > + return 0; > > /* get required size */ > required_size = bpf_read_branch_records(ctx, NULL, 0, > @@ -40,14 +37,14 @@ int perf_branches(void *ctx) > written_global = bpf_read_branch_records(ctx, fpbe, sizeof(fpbe), 0); > /* ignore spurious events */ > if (!written_global) > - return 1; > + return 0; > > required_size_out = required_size; > written_stack_out = written_stack; > written_global_out = written_global; > - valid = 1; > + __atomic_add_fetch(&valid, 1, __ATOMIC_RELEASE); > > - return 0; > + return 1; > } > > char _license[] SEC("license") = "GPL";