From: Alexandru Elisei <alexandru.elisei@arm.com> To: Ricardo Koller <ricarkol@google.com> Cc: kvm@vger.kernel.org, kvmarm@lists.cs.columbia.edu, drjones@redhat.com, maz@kernel.org, eric.auger@redhat.com, oliver.upton@linux.dev, reijiw@google.com Subject: Re: [kvm-unit-tests PATCH 1/3] arm: pmu: Add missing isb()'s after sys register writing Date: Tue, 19 Jul 2022 12:14:37 +0100 [thread overview] Message-ID: <YtaSDhj2SXEzh8QI@monolith.localdoman> (raw) In-Reply-To: <20220718154910.3923412-2-ricarkol@google.com> Hi, Since you're touching the PMU tests, I took the liberty to suggest changes somewhat related to this patch. If you don't want to implement them, let me know and I'll try to make a patch/series out of them. On Mon, Jul 18, 2022 at 08:49:08AM -0700, Ricardo Koller wrote: > There are various pmu tests that require an isb() between enabling > counting and the actual counting. This can lead to count registers > reporting less events than expected; the actual enabling happens after > some events have happened. For example, some missing isb()'s in the > pmu-sw-incr test lead to the following errors on bare-metal: > > INFO: pmu: pmu-sw-incr: SW_INCR counter #0 has value 4294967280 > PASS: pmu: pmu-sw-incr: PWSYNC does not increment if PMCR.E is unset > FAIL: pmu: pmu-sw-incr: counter #1 after + 100 SW_INCR > FAIL: pmu: pmu-sw-incr: counter #0 after + 100 SW_INCR > INFO: pmu: pmu-sw-incr: counter values after 100 SW_INCR #0=82 #1=98 > PASS: pmu: pmu-sw-incr: overflow on counter #0 after 100 SW_INCR > SUMMARY: 4 tests, 2 unexpected failures > > Add the missing isb()'s on all failing tests, plus some others that are > not currently required but might in the future (like an isb() after > clearing the overflow signal in the IRQ handler). > > Signed-off-by: Ricardo Koller <ricarkol@google.com> > --- > arm/pmu.c | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/arm/pmu.c b/arm/pmu.c > index 15c542a2..fd838392 100644 > --- a/arm/pmu.c > +++ b/arm/pmu.c > @@ -307,6 +307,7 @@ static void irq_handler(struct pt_regs *regs) > } > } > write_sysreg(ALL_SET, pmovsclr_el0); > + isb(); > } else { > pmu_stats.unexpected = true; > } > @@ -534,6 +535,7 @@ static void test_sw_incr(void) > write_sysreg_s(0x3, PMCNTENSET_EL0); > > write_regn_el0(pmevcntr, 0, PRE_OVERFLOW); > + isb(); > > for (i = 0; i < 100; i++) > write_sysreg(0x1, pmswinc_el0); > @@ -547,6 +549,7 @@ static void test_sw_incr(void) > write_regn_el0(pmevcntr, 0, PRE_OVERFLOW); > write_sysreg_s(0x3, PMCNTENSET_EL0); > set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); > + isb(); > > for (i = 0; i < 100; i++) > write_sysreg(0x3, pmswinc_el0); > @@ -618,6 +621,8 @@ static void test_chained_sw_incr(void) > > write_regn_el0(pmevcntr, 0, PRE_OVERFLOW); > set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); > + isb(); > + > for (i = 0; i < 100; i++) > write_sysreg(0x1, pmswinc_el0); > > @@ -634,6 +639,8 @@ static void test_chained_sw_incr(void) > write_regn_el0(pmevcntr, 1, ALL_SET); > write_sysreg_s(0x3, PMCNTENSET_EL0); > set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); > + isb(); > + > for (i = 0; i < 100; i++) > write_sysreg(0x1, pmswinc_el0); > > @@ -821,6 +828,8 @@ static void test_overflow_interrupt(void) > report(expect_interrupts(0), "no overflow interrupt after preset"); > > set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); > + isb(); > + > for (i = 0; i < 100; i++) > write_sysreg(0x2, pmswinc_el0); You missed the set_pmcr(pmu.pmcr_ro) call on the next line. Also the comment "enable interrupts" below: [..] report(expect_interrupts(0), "no overflow interrupt after preset"); set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); for (i = 0; i < 100; i++) write_sysreg(0x2, pmswinc_el0); set_pmcr(pmu.pmcr_ro); report(expect_interrupts(0), "no overflow interrupt after counting"); /* enable interrupts */ pmu_reset_stats(); [..] is misleading, because pmu_reset_stats() doesn't enable the PMU. Unless the intention was to call pmu_reset(), in which case the comment is correct and the code is wrong. My guess is that the comment is incorrect, the test seems to be working fine when the PMU is enabled in the mem_access_loop() call. > > @@ -879,6 +888,7 @@ static bool check_cycles_increase(void) > set_pmccfiltr(0); /* count cycles in EL0, EL1, but not EL2 */ > > set_pmcr(get_pmcr() | PMU_PMCR_LC | PMU_PMCR_C | PMU_PMCR_E); > + isb(); > > for (int i = 0; i < NR_SAMPLES; i++) { > uint64_t a, b; > @@ -894,6 +904,7 @@ static bool check_cycles_increase(void) > } > > set_pmcr(get_pmcr() & ~PMU_PMCR_E); > + isb(); Those look good to me. Thanks, Alex > > return success; > } > -- > 2.37.0.170.g444d1eabd0-goog >
WARNING: multiple messages have this Message-ID (diff)
From: Alexandru Elisei <alexandru.elisei@arm.com> To: Ricardo Koller <ricarkol@google.com> Cc: drjones@redhat.com, kvm@vger.kernel.org, maz@kernel.org, oliver.upton@linux.dev, kvmarm@lists.cs.columbia.edu Subject: Re: [kvm-unit-tests PATCH 1/3] arm: pmu: Add missing isb()'s after sys register writing Date: Tue, 19 Jul 2022 12:14:37 +0100 [thread overview] Message-ID: <YtaSDhj2SXEzh8QI@monolith.localdoman> (raw) In-Reply-To: <20220718154910.3923412-2-ricarkol@google.com> Hi, Since you're touching the PMU tests, I took the liberty to suggest changes somewhat related to this patch. If you don't want to implement them, let me know and I'll try to make a patch/series out of them. On Mon, Jul 18, 2022 at 08:49:08AM -0700, Ricardo Koller wrote: > There are various pmu tests that require an isb() between enabling > counting and the actual counting. This can lead to count registers > reporting less events than expected; the actual enabling happens after > some events have happened. For example, some missing isb()'s in the > pmu-sw-incr test lead to the following errors on bare-metal: > > INFO: pmu: pmu-sw-incr: SW_INCR counter #0 has value 4294967280 > PASS: pmu: pmu-sw-incr: PWSYNC does not increment if PMCR.E is unset > FAIL: pmu: pmu-sw-incr: counter #1 after + 100 SW_INCR > FAIL: pmu: pmu-sw-incr: counter #0 after + 100 SW_INCR > INFO: pmu: pmu-sw-incr: counter values after 100 SW_INCR #0=82 #1=98 > PASS: pmu: pmu-sw-incr: overflow on counter #0 after 100 SW_INCR > SUMMARY: 4 tests, 2 unexpected failures > > Add the missing isb()'s on all failing tests, plus some others that are > not currently required but might in the future (like an isb() after > clearing the overflow signal in the IRQ handler). > > Signed-off-by: Ricardo Koller <ricarkol@google.com> > --- > arm/pmu.c | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/arm/pmu.c b/arm/pmu.c > index 15c542a2..fd838392 100644 > --- a/arm/pmu.c > +++ b/arm/pmu.c > @@ -307,6 +307,7 @@ static void irq_handler(struct pt_regs *regs) > } > } > write_sysreg(ALL_SET, pmovsclr_el0); > + isb(); > } else { > pmu_stats.unexpected = true; > } > @@ -534,6 +535,7 @@ static void test_sw_incr(void) > write_sysreg_s(0x3, PMCNTENSET_EL0); > > write_regn_el0(pmevcntr, 0, PRE_OVERFLOW); > + isb(); > > for (i = 0; i < 100; i++) > write_sysreg(0x1, pmswinc_el0); > @@ -547,6 +549,7 @@ static void test_sw_incr(void) > write_regn_el0(pmevcntr, 0, PRE_OVERFLOW); > write_sysreg_s(0x3, PMCNTENSET_EL0); > set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); > + isb(); > > for (i = 0; i < 100; i++) > write_sysreg(0x3, pmswinc_el0); > @@ -618,6 +621,8 @@ static void test_chained_sw_incr(void) > > write_regn_el0(pmevcntr, 0, PRE_OVERFLOW); > set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); > + isb(); > + > for (i = 0; i < 100; i++) > write_sysreg(0x1, pmswinc_el0); > > @@ -634,6 +639,8 @@ static void test_chained_sw_incr(void) > write_regn_el0(pmevcntr, 1, ALL_SET); > write_sysreg_s(0x3, PMCNTENSET_EL0); > set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); > + isb(); > + > for (i = 0; i < 100; i++) > write_sysreg(0x1, pmswinc_el0); > > @@ -821,6 +828,8 @@ static void test_overflow_interrupt(void) > report(expect_interrupts(0), "no overflow interrupt after preset"); > > set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); > + isb(); > + > for (i = 0; i < 100; i++) > write_sysreg(0x2, pmswinc_el0); You missed the set_pmcr(pmu.pmcr_ro) call on the next line. Also the comment "enable interrupts" below: [..] report(expect_interrupts(0), "no overflow interrupt after preset"); set_pmcr(pmu.pmcr_ro | PMU_PMCR_E); for (i = 0; i < 100; i++) write_sysreg(0x2, pmswinc_el0); set_pmcr(pmu.pmcr_ro); report(expect_interrupts(0), "no overflow interrupt after counting"); /* enable interrupts */ pmu_reset_stats(); [..] is misleading, because pmu_reset_stats() doesn't enable the PMU. Unless the intention was to call pmu_reset(), in which case the comment is correct and the code is wrong. My guess is that the comment is incorrect, the test seems to be working fine when the PMU is enabled in the mem_access_loop() call. > > @@ -879,6 +888,7 @@ static bool check_cycles_increase(void) > set_pmccfiltr(0); /* count cycles in EL0, EL1, but not EL2 */ > > set_pmcr(get_pmcr() | PMU_PMCR_LC | PMU_PMCR_C | PMU_PMCR_E); > + isb(); > > for (int i = 0; i < NR_SAMPLES; i++) { > uint64_t a, b; > @@ -894,6 +904,7 @@ static bool check_cycles_increase(void) > } > > set_pmcr(get_pmcr() & ~PMU_PMCR_E); > + isb(); Those look good to me. Thanks, Alex > > return success; > } > -- > 2.37.0.170.g444d1eabd0-goog > _______________________________________________ kvmarm mailing list kvmarm@lists.cs.columbia.edu https://lists.cs.columbia.edu/mailman/listinfo/kvmarm
next prev parent reply other threads:[~2022-07-19 11:14 UTC|newest] Thread overview: 48+ messages / expand[flat|nested] mbox.gz Atom feed top 2022-07-18 15:49 [kvm-unit-tests PATCH 0/3] arm: pmu: Fixes for bare metal Ricardo Koller 2022-07-18 15:49 ` Ricardo Koller 2022-07-18 15:49 ` [kvm-unit-tests PATCH 1/3] arm: pmu: Add missing isb()'s after sys register writing Ricardo Koller 2022-07-18 15:49 ` Ricardo Koller 2022-07-18 16:38 ` Alexandru Elisei 2022-07-18 16:38 ` Alexandru Elisei 2022-07-18 17:48 ` Ricardo Koller 2022-07-18 17:48 ` Ricardo Koller 2022-07-19 11:26 ` Alexandru Elisei 2022-07-19 11:26 ` Alexandru Elisei 2022-07-19 11:14 ` Alexandru Elisei [this message] 2022-07-19 11:14 ` Alexandru Elisei 2022-07-20 21:20 ` Ricardo Koller 2022-07-20 21:20 ` Ricardo Koller 2022-07-18 15:49 ` [kvm-unit-tests PATCH 2/3] arm: pmu: Reset the pmu registers before starting some tests Ricardo Koller 2022-07-18 15:49 ` Ricardo Koller 2022-07-18 15:49 ` [kvm-unit-tests PATCH 3/3] arm: pmu: Remove checks for !overflow in chained counters tests Ricardo Koller 2022-07-18 15:49 ` Ricardo Koller 2022-07-19 11:34 ` Marc Zyngier 2022-07-19 11:34 ` Marc Zyngier 2022-07-20 8:40 ` Ricardo Koller 2022-07-20 8:40 ` Ricardo Koller 2022-07-20 9:45 ` Marc Zyngier 2022-07-20 9:45 ` Marc Zyngier 2022-07-20 21:17 ` Ricardo Koller 2022-07-20 21:17 ` Ricardo Koller 2022-07-20 21:26 ` Ricardo Koller 2022-07-20 21:26 ` Ricardo Koller 2022-07-21 13:43 ` Marc Zyngier 2022-07-21 13:43 ` Marc Zyngier 2022-07-22 21:53 ` Ricardo Koller 2022-07-22 21:53 ` Ricardo Koller 2022-07-23 7:59 ` Andrew Jones 2022-07-23 7:59 ` Andrew Jones 2022-07-24 9:40 ` Marc Zyngier 2022-07-24 9:40 ` Marc Zyngier 2022-07-27 2:29 ` Ricardo Koller 2022-07-27 2:29 ` Ricardo Koller 2022-07-30 12:47 ` Marc Zyngier 2022-07-30 12:47 ` Marc Zyngier 2022-07-30 12:52 ` Marc Zyngier 2022-07-30 12:52 ` Marc Zyngier 2022-08-01 19:15 ` Ricardo Koller 2022-08-01 19:15 ` Ricardo Koller 2022-07-18 16:42 ` [kvm-unit-tests PATCH 0/3] arm: pmu: Fixes for bare metal Alexandru Elisei 2022-07-18 16:42 ` Alexandru Elisei 2022-07-18 17:18 ` Ricardo Koller 2022-07-18 17:18 ` Ricardo Koller
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=YtaSDhj2SXEzh8QI@monolith.localdoman \ --to=alexandru.elisei@arm.com \ --cc=drjones@redhat.com \ --cc=eric.auger@redhat.com \ --cc=kvm@vger.kernel.org \ --cc=kvmarm@lists.cs.columbia.edu \ --cc=maz@kernel.org \ --cc=oliver.upton@linux.dev \ --cc=reijiw@google.com \ --cc=ricarkol@google.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: linkBe sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.