* [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-22 16:03 ` Phil Auld 0 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-22 16:03 UTC (permalink / raw) To: linux-kernel Cc: Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel Some arm64 rely on store_cpu_topology() to setup the real topology. This needs to be done before the call to notify_cpu_starting() which tell the scheduler about the cpu otherwise the core scheduling data structures are setup in a way that does not match the actual topology. Without this change stress-ng (which enables core scheduling in its prctl tests) causes a warning and then a crash (trimmed for legibility): [ 1853.805168] ------------[ cut here ]------------ [ 1853.809784] task_rq(b)->core != rq->core [ 1853.809792] WARNING: CPU: 117 PID: 0 at kernel/sched/fair.c:11102 cfs_prio_less+0x1b4/0x1c4 ... [ 1854.015210] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010 ... [ 1854.231256] Call trace: [ 1854.233689] pick_next_task+0x3dc/0x81c [ 1854.237512] __schedule+0x10c/0x4cc [ 1854.240988] schedule_idle+0x34/0x54 Fixes: 9edeaea1bc45 ("sched: Core-wide rq->lock") Signed-off-by: Phil Auld <pauld@redhat.com> --- This is a similar issue to f2703def339c ("MIPS: smp: fill in sibling and core maps earlier") which fixed it for MIPS. arch/arm64/kernel/smp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c index 27df5c1e6baa..3b46041f2b97 100644 --- a/arch/arm64/kernel/smp.c +++ b/arch/arm64/kernel/smp.c @@ -234,6 +234,7 @@ asmlinkage notrace void secondary_start_kernel(void) * Log the CPU info before it is marked online and might get read. */ cpuinfo_store_cpu(); + store_cpu_topology(cpu); /* * Enable GIC and timers. @@ -242,7 +243,6 @@ asmlinkage notrace void secondary_start_kernel(void) ipi_setup(cpu); - store_cpu_topology(cpu); numa_add_cpu(cpu); /* -- 2.18.0 ^ permalink raw reply related [flat|nested] 16+ messages in thread
* [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-22 16:03 ` Phil Auld 0 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-22 16:03 UTC (permalink / raw) To: linux-kernel Cc: Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel Some arm64 rely on store_cpu_topology() to setup the real topology. This needs to be done before the call to notify_cpu_starting() which tell the scheduler about the cpu otherwise the core scheduling data structures are setup in a way that does not match the actual topology. Without this change stress-ng (which enables core scheduling in its prctl tests) causes a warning and then a crash (trimmed for legibility): [ 1853.805168] ------------[ cut here ]------------ [ 1853.809784] task_rq(b)->core != rq->core [ 1853.809792] WARNING: CPU: 117 PID: 0 at kernel/sched/fair.c:11102 cfs_prio_less+0x1b4/0x1c4 ... [ 1854.015210] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010 ... [ 1854.231256] Call trace: [ 1854.233689] pick_next_task+0x3dc/0x81c [ 1854.237512] __schedule+0x10c/0x4cc [ 1854.240988] schedule_idle+0x34/0x54 Fixes: 9edeaea1bc45 ("sched: Core-wide rq->lock") Signed-off-by: Phil Auld <pauld@redhat.com> --- This is a similar issue to f2703def339c ("MIPS: smp: fill in sibling and core maps earlier") which fixed it for MIPS. arch/arm64/kernel/smp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c index 27df5c1e6baa..3b46041f2b97 100644 --- a/arch/arm64/kernel/smp.c +++ b/arch/arm64/kernel/smp.c @@ -234,6 +234,7 @@ asmlinkage notrace void secondary_start_kernel(void) * Log the CPU info before it is marked online and might get read. */ cpuinfo_store_cpu(); + store_cpu_topology(cpu); /* * Enable GIC and timers. @@ -242,7 +243,6 @@ asmlinkage notrace void secondary_start_kernel(void) ipi_setup(cpu); - store_cpu_topology(cpu); numa_add_cpu(cpu); /* -- 2.18.0 _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply related [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling 2022-03-22 16:03 ` Phil Auld @ 2022-03-29 14:02 ` Dietmar Eggemann -1 siblings, 0 replies; 16+ messages in thread From: Dietmar Eggemann @ 2022-03-29 14:02 UTC (permalink / raw) To: Phil Auld, linux-kernel Cc: Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On 22/03/2022 17:03, Phil Auld wrote: > Some arm64 rely on store_cpu_topology() to setup the real topology. > This needs to be done before the call to notify_cpu_starting() which > tell the scheduler about the cpu otherwise the core scheduling data > structures are setup in a way that does not match the actual topology. > > Without this change stress-ng (which enables core scheduling in its prctl > tests) causes a warning and then a crash (trimmed for legibility): > > [ 1853.805168] ------------[ cut here ]------------ > [ 1853.809784] task_rq(b)->core != rq->core > [ 1853.809792] WARNING: CPU: 117 PID: 0 at kernel/sched/fair.c:11102 cfs_prio_less+0x1b4/0x1c4 > ... > [ 1854.015210] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010 > ... > [ 1854.231256] Call trace: > [ 1854.233689] pick_next_task+0x3dc/0x81c > [ 1854.237512] __schedule+0x10c/0x4cc > [ 1854.240988] schedule_idle+0x34/0x54 > > Fixes: 9edeaea1bc45 ("sched: Core-wide rq->lock") > Signed-off-by: Phil Auld <pauld@redhat.com> > --- > This is a similar issue to > f2703def339c ("MIPS: smp: fill in sibling and core maps earlier") > which fixed it for MIPS. I assume this is for a machine which relies on MPIDR-based setup (package_id == -1)? I.e. it doesn't have proper ACPI/(DT) data for topology setup. Tried on a ThunderX2 by disabling parse_acpi_topology() but then I end up with a machine w/o SMT, so `stress-ng --prctl N` doesn't show this issue. Which machine were you using? > arch/arm64/kernel/smp.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c > index 27df5c1e6baa..3b46041f2b97 100644 > --- a/arch/arm64/kernel/smp.c > +++ b/arch/arm64/kernel/smp.c > @@ -234,6 +234,7 @@ asmlinkage notrace void secondary_start_kernel(void) > * Log the CPU info before it is marked online and might get read. > */ > cpuinfo_store_cpu(); > + store_cpu_topology(cpu); > > /* > * Enable GIC and timers. > @@ -242,7 +243,6 @@ asmlinkage notrace void secondary_start_kernel(void) > > ipi_setup(cpu); > > - store_cpu_topology(cpu); > numa_add_cpu(cpu); > > /* ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-29 14:02 ` Dietmar Eggemann 0 siblings, 0 replies; 16+ messages in thread From: Dietmar Eggemann @ 2022-03-29 14:02 UTC (permalink / raw) To: Phil Auld, linux-kernel Cc: Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On 22/03/2022 17:03, Phil Auld wrote: > Some arm64 rely on store_cpu_topology() to setup the real topology. > This needs to be done before the call to notify_cpu_starting() which > tell the scheduler about the cpu otherwise the core scheduling data > structures are setup in a way that does not match the actual topology. > > Without this change stress-ng (which enables core scheduling in its prctl > tests) causes a warning and then a crash (trimmed for legibility): > > [ 1853.805168] ------------[ cut here ]------------ > [ 1853.809784] task_rq(b)->core != rq->core > [ 1853.809792] WARNING: CPU: 117 PID: 0 at kernel/sched/fair.c:11102 cfs_prio_less+0x1b4/0x1c4 > ... > [ 1854.015210] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010 > ... > [ 1854.231256] Call trace: > [ 1854.233689] pick_next_task+0x3dc/0x81c > [ 1854.237512] __schedule+0x10c/0x4cc > [ 1854.240988] schedule_idle+0x34/0x54 > > Fixes: 9edeaea1bc45 ("sched: Core-wide rq->lock") > Signed-off-by: Phil Auld <pauld@redhat.com> > --- > This is a similar issue to > f2703def339c ("MIPS: smp: fill in sibling and core maps earlier") > which fixed it for MIPS. I assume this is for a machine which relies on MPIDR-based setup (package_id == -1)? I.e. it doesn't have proper ACPI/(DT) data for topology setup. Tried on a ThunderX2 by disabling parse_acpi_topology() but then I end up with a machine w/o SMT, so `stress-ng --prctl N` doesn't show this issue. Which machine were you using? > arch/arm64/kernel/smp.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c > index 27df5c1e6baa..3b46041f2b97 100644 > --- a/arch/arm64/kernel/smp.c > +++ b/arch/arm64/kernel/smp.c > @@ -234,6 +234,7 @@ asmlinkage notrace void secondary_start_kernel(void) > * Log the CPU info before it is marked online and might get read. > */ > cpuinfo_store_cpu(); > + store_cpu_topology(cpu); > > /* > * Enable GIC and timers. > @@ -242,7 +243,6 @@ asmlinkage notrace void secondary_start_kernel(void) > > ipi_setup(cpu); > > - store_cpu_topology(cpu); > numa_add_cpu(cpu); > > /* _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling 2022-03-29 14:02 ` Dietmar Eggemann @ 2022-03-29 15:20 ` Phil Auld -1 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-29 15:20 UTC (permalink / raw) To: Dietmar Eggemann Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: > On 22/03/2022 17:03, Phil Auld wrote: > > Some arm64 rely on store_cpu_topology() to setup the real topology. > > This needs to be done before the call to notify_cpu_starting() which > > tell the scheduler about the cpu otherwise the core scheduling data > > structures are setup in a way that does not match the actual topology. > > > > Without this change stress-ng (which enables core scheduling in its prctl > > tests) causes a warning and then a crash (trimmed for legibility): > > > > [ 1853.805168] ------------[ cut here ]------------ > > [ 1853.809784] task_rq(b)->core != rq->core > > [ 1853.809792] WARNING: CPU: 117 PID: 0 at kernel/sched/fair.c:11102 cfs_prio_less+0x1b4/0x1c4 > > ... > > [ 1854.015210] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010 > > ... > > [ 1854.231256] Call trace: > > [ 1854.233689] pick_next_task+0x3dc/0x81c > > [ 1854.237512] __schedule+0x10c/0x4cc > > [ 1854.240988] schedule_idle+0x34/0x54 > > > > Fixes: 9edeaea1bc45 ("sched: Core-wide rq->lock") > > Signed-off-by: Phil Auld <pauld@redhat.com> > > --- > > This is a similar issue to > > f2703def339c ("MIPS: smp: fill in sibling and core maps earlier") > > which fixed it for MIPS. > > I assume this is for a machine which relies on MPIDR-based setup > (package_id == -1)? I.e. it doesn't have proper ACPI/(DT) data for > topology setup. Yes, that's my understanding. No PPTT. > > Tried on a ThunderX2 by disabling parse_acpi_topology() but then I end > up with a machine w/o SMT, so `stress-ng --prctl N` doesn't show this issue. > > Which machine were you using? This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 chips. ARM (CN9980-2200LG4077-Y21-G) Thanks, Phil > > > arch/arm64/kernel/smp.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c > > index 27df5c1e6baa..3b46041f2b97 100644 > > --- a/arch/arm64/kernel/smp.c > > +++ b/arch/arm64/kernel/smp.c > > @@ -234,6 +234,7 @@ asmlinkage notrace void secondary_start_kernel(void) > > * Log the CPU info before it is marked online and might get read. > > */ > > cpuinfo_store_cpu(); > > + store_cpu_topology(cpu); > > > > /* > > * Enable GIC and timers. > > @@ -242,7 +243,6 @@ asmlinkage notrace void secondary_start_kernel(void) > > > > ipi_setup(cpu); > > > > - store_cpu_topology(cpu); > > numa_add_cpu(cpu); > > > > /* > -- ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-29 15:20 ` Phil Auld 0 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-29 15:20 UTC (permalink / raw) To: Dietmar Eggemann Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: > On 22/03/2022 17:03, Phil Auld wrote: > > Some arm64 rely on store_cpu_topology() to setup the real topology. > > This needs to be done before the call to notify_cpu_starting() which > > tell the scheduler about the cpu otherwise the core scheduling data > > structures are setup in a way that does not match the actual topology. > > > > Without this change stress-ng (which enables core scheduling in its prctl > > tests) causes a warning and then a crash (trimmed for legibility): > > > > [ 1853.805168] ------------[ cut here ]------------ > > [ 1853.809784] task_rq(b)->core != rq->core > > [ 1853.809792] WARNING: CPU: 117 PID: 0 at kernel/sched/fair.c:11102 cfs_prio_less+0x1b4/0x1c4 > > ... > > [ 1854.015210] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010 > > ... > > [ 1854.231256] Call trace: > > [ 1854.233689] pick_next_task+0x3dc/0x81c > > [ 1854.237512] __schedule+0x10c/0x4cc > > [ 1854.240988] schedule_idle+0x34/0x54 > > > > Fixes: 9edeaea1bc45 ("sched: Core-wide rq->lock") > > Signed-off-by: Phil Auld <pauld@redhat.com> > > --- > > This is a similar issue to > > f2703def339c ("MIPS: smp: fill in sibling and core maps earlier") > > which fixed it for MIPS. > > I assume this is for a machine which relies on MPIDR-based setup > (package_id == -1)? I.e. it doesn't have proper ACPI/(DT) data for > topology setup. Yes, that's my understanding. No PPTT. > > Tried on a ThunderX2 by disabling parse_acpi_topology() but then I end > up with a machine w/o SMT, so `stress-ng --prctl N` doesn't show this issue. > > Which machine were you using? This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 chips. ARM (CN9980-2200LG4077-Y21-G) Thanks, Phil > > > arch/arm64/kernel/smp.c | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/arch/arm64/kernel/smp.c b/arch/arm64/kernel/smp.c > > index 27df5c1e6baa..3b46041f2b97 100644 > > --- a/arch/arm64/kernel/smp.c > > +++ b/arch/arm64/kernel/smp.c > > @@ -234,6 +234,7 @@ asmlinkage notrace void secondary_start_kernel(void) > > * Log the CPU info before it is marked online and might get read. > > */ > > cpuinfo_store_cpu(); > > + store_cpu_topology(cpu); > > > > /* > > * Enable GIC and timers. > > @@ -242,7 +243,6 @@ asmlinkage notrace void secondary_start_kernel(void) > > > > ipi_setup(cpu); > > > > - store_cpu_topology(cpu); > > numa_add_cpu(cpu); > > > > /* > -- _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling 2022-03-29 15:20 ` Phil Auld @ 2022-03-29 18:55 ` Dietmar Eggemann -1 siblings, 0 replies; 16+ messages in thread From: Dietmar Eggemann @ 2022-03-29 18:55 UTC (permalink / raw) To: Phil Auld Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On 29/03/2022 17:20, Phil Auld wrote: > On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: >> On 22/03/2022 17:03, Phil Auld wrote: [...] >> I assume this is for a machine which relies on MPIDR-based setup >> (package_id == -1)? I.e. it doesn't have proper ACPI/(DT) data for >> topology setup. > > Yes, that's my understanding. No PPTT. > >> >> Tried on a ThunderX2 by disabling parse_acpi_topology() but then I end >> up with a machine w/o SMT, so `stress-ng --prctl N` doesn't show this issue. >> >> Which machine were you using? > > This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 > chips. > > ARM (CN9980-2200LG4077-Y21-G) I'm using the same processor just with ACPI/PPTT. # sudo dmidecode -t 4 | grep "Part Number" Part Number: CN9980-2200LG4077-21-Y-G Part Number: CN9980-2200LG4077-21-Y-G # cat /sys/devices/system/cpu/cpu0/topology/thread_siblings 0,32,64,96 # cat /sys/kernel/debug/sched/domains/cpu0/domain*/name SMT MC NUMA But no matter whether I disable parse_acpi_topology() or just force `cpu_topology[cpu].package_id = -1` in this function, I always end up with: # cat /sys/kernel/debug/sched/domains/cpu0/domain*/name MC NUMA # cat /sys/devices/system/cpu/cpu0/topology/thread_siblings_list 0 so no SMT sched domain. The MPIDR-based topology fallback code in store_cpu_topology() forces `cpuid_topo->thread_id = -1`. IMHO this is why on my machine I don't see this issue while running: root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 stress-ng: info: [2388042] dispatching hogs: 256 prctl Is there something I miss in my setup to provoke this issue? [...] ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-29 18:55 ` Dietmar Eggemann 0 siblings, 0 replies; 16+ messages in thread From: Dietmar Eggemann @ 2022-03-29 18:55 UTC (permalink / raw) To: Phil Auld Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On 29/03/2022 17:20, Phil Auld wrote: > On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: >> On 22/03/2022 17:03, Phil Auld wrote: [...] >> I assume this is for a machine which relies on MPIDR-based setup >> (package_id == -1)? I.e. it doesn't have proper ACPI/(DT) data for >> topology setup. > > Yes, that's my understanding. No PPTT. > >> >> Tried on a ThunderX2 by disabling parse_acpi_topology() but then I end >> up with a machine w/o SMT, so `stress-ng --prctl N` doesn't show this issue. >> >> Which machine were you using? > > This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 > chips. > > ARM (CN9980-2200LG4077-Y21-G) I'm using the same processor just with ACPI/PPTT. # sudo dmidecode -t 4 | grep "Part Number" Part Number: CN9980-2200LG4077-21-Y-G Part Number: CN9980-2200LG4077-21-Y-G # cat /sys/devices/system/cpu/cpu0/topology/thread_siblings 0,32,64,96 # cat /sys/kernel/debug/sched/domains/cpu0/domain*/name SMT MC NUMA But no matter whether I disable parse_acpi_topology() or just force `cpu_topology[cpu].package_id = -1` in this function, I always end up with: # cat /sys/kernel/debug/sched/domains/cpu0/domain*/name MC NUMA # cat /sys/devices/system/cpu/cpu0/topology/thread_siblings_list 0 so no SMT sched domain. The MPIDR-based topology fallback code in store_cpu_topology() forces `cpuid_topo->thread_id = -1`. IMHO this is why on my machine I don't see this issue while running: root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 stress-ng: info: [2388042] dispatching hogs: 256 prctl Is there something I miss in my setup to provoke this issue? [...] _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling 2022-03-29 18:55 ` Dietmar Eggemann @ 2022-03-29 19:50 ` Phil Auld -1 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-29 19:50 UTC (permalink / raw) To: Dietmar Eggemann Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On Tue, Mar 29, 2022 at 08:55:08PM +0200 Dietmar Eggemann wrote: > On 29/03/2022 17:20, Phil Auld wrote: > > On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: > >> On 22/03/2022 17:03, Phil Auld wrote: > > [...] > > >> I assume this is for a machine which relies on MPIDR-based setup > >> (package_id == -1)? I.e. it doesn't have proper ACPI/(DT) data for > >> topology setup. > > > > Yes, that's my understanding. No PPTT. > > > >> > >> Tried on a ThunderX2 by disabling parse_acpi_topology() but then I end > >> up with a machine w/o SMT, so `stress-ng --prctl N` doesn't show this issue. > >> > >> Which machine were you using? > > > > This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 > > chips. > > > > ARM (CN9980-2200LG4077-Y21-G) > I'm using the same processor just with ACPI/PPTT. > Maybe I'm misinformed about these systems having no PPTT... I'm reclaiming the system. Is there a way I can tell from userspace? > # sudo dmidecode -t 4 | grep "Part Number" > Part Number: CN9980-2200LG4077-21-Y-G > Part Number: CN9980-2200LG4077-21-Y-G > > # cat /sys/devices/system/cpu/cpu0/topology/thread_siblings > 0,32,64,96 > > # cat /sys/kernel/debug/sched/domains/cpu0/domain*/name > SMT > MC > NUMA > > But no matter whether I disable parse_acpi_topology() or just force > `cpu_topology[cpu].package_id = -1` in this function, I always end up with: > > # cat /sys/kernel/debug/sched/domains/cpu0/domain*/name > MC > NUMA > > # cat /sys/devices/system/cpu/cpu0/topology/thread_siblings_list > 0 > > so no SMT sched domain. The MPIDR-based topology fallback code in > store_cpu_topology() forces `cpuid_topo->thread_id = -1`. Right. So since I'm getting SMT it must not have package_id == -1. In which case you should be able to reproduce it because it must be that the call the update_siblings_masks() is required. That appears to only be called from store_cpu_topology() which is after the scheduler has already setup the core pointers. The fix could be the same but I should reword the commit message since it should effect all SMT arm systems I'd think. Or maybe the ACPI topology code should call update_sibling_masks(). > > IMHO this is why on my machine I don't see this issue while running: > > root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 > stress-ng: info: [2388042] dispatching hogs: 256 prctl > > Is there something I miss in my setup to provoke this issue? > Make sure you have a stress-ng that is new enough and built against headers that have the CORE_SCHED prctls defined. BTW, thanks for taking a look. Cheers, Phil > [...] > -- ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-29 19:50 ` Phil Auld 0 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-29 19:50 UTC (permalink / raw) To: Dietmar Eggemann Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On Tue, Mar 29, 2022 at 08:55:08PM +0200 Dietmar Eggemann wrote: > On 29/03/2022 17:20, Phil Auld wrote: > > On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: > >> On 22/03/2022 17:03, Phil Auld wrote: > > [...] > > >> I assume this is for a machine which relies on MPIDR-based setup > >> (package_id == -1)? I.e. it doesn't have proper ACPI/(DT) data for > >> topology setup. > > > > Yes, that's my understanding. No PPTT. > > > >> > >> Tried on a ThunderX2 by disabling parse_acpi_topology() but then I end > >> up with a machine w/o SMT, so `stress-ng --prctl N` doesn't show this issue. > >> > >> Which machine were you using? > > > > This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 > > chips. > > > > ARM (CN9980-2200LG4077-Y21-G) > I'm using the same processor just with ACPI/PPTT. > Maybe I'm misinformed about these systems having no PPTT... I'm reclaiming the system. Is there a way I can tell from userspace? > # sudo dmidecode -t 4 | grep "Part Number" > Part Number: CN9980-2200LG4077-21-Y-G > Part Number: CN9980-2200LG4077-21-Y-G > > # cat /sys/devices/system/cpu/cpu0/topology/thread_siblings > 0,32,64,96 > > # cat /sys/kernel/debug/sched/domains/cpu0/domain*/name > SMT > MC > NUMA > > But no matter whether I disable parse_acpi_topology() or just force > `cpu_topology[cpu].package_id = -1` in this function, I always end up with: > > # cat /sys/kernel/debug/sched/domains/cpu0/domain*/name > MC > NUMA > > # cat /sys/devices/system/cpu/cpu0/topology/thread_siblings_list > 0 > > so no SMT sched domain. The MPIDR-based topology fallback code in > store_cpu_topology() forces `cpuid_topo->thread_id = -1`. Right. So since I'm getting SMT it must not have package_id == -1. In which case you should be able to reproduce it because it must be that the call the update_siblings_masks() is required. That appears to only be called from store_cpu_topology() which is after the scheduler has already setup the core pointers. The fix could be the same but I should reword the commit message since it should effect all SMT arm systems I'd think. Or maybe the ACPI topology code should call update_sibling_masks(). > > IMHO this is why on my machine I don't see this issue while running: > > root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 > stress-ng: info: [2388042] dispatching hogs: 256 prctl > > Is there something I miss in my setup to provoke this issue? > Make sure you have a stress-ng that is new enough and built against headers that have the CORE_SCHED prctls defined. BTW, thanks for taking a look. Cheers, Phil > [...] > -- _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling 2022-03-29 19:50 ` Phil Auld @ 2022-03-30 15:48 ` Dietmar Eggemann -1 siblings, 0 replies; 16+ messages in thread From: Dietmar Eggemann @ 2022-03-30 15:48 UTC (permalink / raw) To: Phil Auld Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On 29/03/2022 21:50, Phil Auld wrote: > On Tue, Mar 29, 2022 at 08:55:08PM +0200 Dietmar Eggemann wrote: >> On 29/03/2022 17:20, Phil Auld wrote: >>> On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: >>>> On 22/03/2022 17:03, Phil Auld wrote: [...] >>> This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 >>> chips. >>> >>> ARM (CN9980-2200LG4077-Y21-G) >> I'm using the same processor just with ACPI/PPTT. >> > > Maybe I'm misinformed about these systems having no PPTT... > > I'm reclaiming the system. Is there a way I can tell from userspace? # cat /sys/firmware/acpi/tables/PPTT > pptt.dat # iasl -d pptt.dat # vim pptt.dsl [...] >> so no SMT sched domain. The MPIDR-based topology fallback code in >> store_cpu_topology() forces `cpuid_topo->thread_id = -1`. > > Right. So since I'm getting SMT it must not have package_id == -1. > In which case you should be able to reproduce it because it must > be that the call the update_siblings_masks() is required. That > appears to only be called from store_cpu_topology() which is > after the scheduler has already setup the core pointers. > > The fix could be the same but I should reword the commit message > since it should effect all SMT arm systems I'd think. > > Or maybe the ACPI topology code should call update_sibling_masks(). >> >> IMHO this is why on my machine I don't see this issue while running: >> >> root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 >> stress-ng: info: [2388042] dispatching hogs: 256 prctl >> >> Is there something I miss in my setup to provoke this issue? >> > > Make sure you have a stress-ng that is new enough and built against > headers that have the CORE_SCHED prctls defined. Ah, I was using a pretty old version 0.11.07. Now I switched to 0.13.12 which includes: 9038e442b92d - stress-prctl: add Linux 5.14 PR_SCHED_CORE prctl To get SCHED_CORE activated in stress-prctl.c, as a quick hack, I had to add the definitions of PR_SCHED_CORE, PR_SCHED_CORE_GET, etc. to this file. Now the issue you described triggers on this machine immediately. ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-30 15:48 ` Dietmar Eggemann 0 siblings, 0 replies; 16+ messages in thread From: Dietmar Eggemann @ 2022-03-30 15:48 UTC (permalink / raw) To: Phil Auld Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On 29/03/2022 21:50, Phil Auld wrote: > On Tue, Mar 29, 2022 at 08:55:08PM +0200 Dietmar Eggemann wrote: >> On 29/03/2022 17:20, Phil Auld wrote: >>> On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: >>>> On 22/03/2022 17:03, Phil Auld wrote: [...] >>> This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 >>> chips. >>> >>> ARM (CN9980-2200LG4077-Y21-G) >> I'm using the same processor just with ACPI/PPTT. >> > > Maybe I'm misinformed about these systems having no PPTT... > > I'm reclaiming the system. Is there a way I can tell from userspace? # cat /sys/firmware/acpi/tables/PPTT > pptt.dat # iasl -d pptt.dat # vim pptt.dsl [...] >> so no SMT sched domain. The MPIDR-based topology fallback code in >> store_cpu_topology() forces `cpuid_topo->thread_id = -1`. > > Right. So since I'm getting SMT it must not have package_id == -1. > In which case you should be able to reproduce it because it must > be that the call the update_siblings_masks() is required. That > appears to only be called from store_cpu_topology() which is > after the scheduler has already setup the core pointers. > > The fix could be the same but I should reword the commit message > since it should effect all SMT arm systems I'd think. > > Or maybe the ACPI topology code should call update_sibling_masks(). >> >> IMHO this is why on my machine I don't see this issue while running: >> >> root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 >> stress-ng: info: [2388042] dispatching hogs: 256 prctl >> >> Is there something I miss in my setup to provoke this issue? >> > > Make sure you have a stress-ng that is new enough and built against > headers that have the CORE_SCHED prctls defined. Ah, I was using a pretty old version 0.11.07. Now I switched to 0.13.12 which includes: 9038e442b92d - stress-prctl: add Linux 5.14 PR_SCHED_CORE prctl To get SCHED_CORE activated in stress-prctl.c, as a quick hack, I had to add the definitions of PR_SCHED_CORE, PR_SCHED_CORE_GET, etc. to this file. Now the issue you described triggers on this machine immediately. _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling 2022-03-30 15:48 ` Dietmar Eggemann @ 2022-03-30 15:52 ` Phil Auld -1 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-30 15:52 UTC (permalink / raw) To: Dietmar Eggemann Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On Wed, Mar 30, 2022 at 05:48:34PM +0200 Dietmar Eggemann wrote: > On 29/03/2022 21:50, Phil Auld wrote: > > On Tue, Mar 29, 2022 at 08:55:08PM +0200 Dietmar Eggemann wrote: > >> On 29/03/2022 17:20, Phil Auld wrote: > >>> On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: > >>>> On 22/03/2022 17:03, Phil Auld wrote: > > [...] > > >>> This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 > >>> chips. > >>> > >>> ARM (CN9980-2200LG4077-Y21-G) > >> I'm using the same processor just with ACPI/PPTT. > >> > > > > Maybe I'm misinformed about these systems having no PPTT... > > > > I'm reclaiming the system. Is there a way I can tell from userspace? > > # cat /sys/firmware/acpi/tables/PPTT > pptt.dat > # iasl -d pptt.dat > # vim pptt.dsl > Thanks, I'll git that a try. I suspect these are the same as yours though and I was just mistaken :) > [...] > > >> so no SMT sched domain. The MPIDR-based topology fallback code in > >> store_cpu_topology() forces `cpuid_topo->thread_id = -1`. > > > > Right. So since I'm getting SMT it must not have package_id == -1. > > In which case you should be able to reproduce it because it must > > be that the call the update_siblings_masks() is required. That > > appears to only be called from store_cpu_topology() which is > > after the scheduler has already setup the core pointers. > > > > The fix could be the same but I should reword the commit message > > since it should effect all SMT arm systems I'd think. > > > > Or maybe the ACPI topology code should call update_sibling_masks(). > >> > >> IMHO this is why on my machine I don't see this issue while running: > >> > >> root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 > >> stress-ng: info: [2388042] dispatching hogs: 256 prctl > >> > >> Is there something I miss in my setup to provoke this issue? > >> > > > > Make sure you have a stress-ng that is new enough and built against > > headers that have the CORE_SCHED prctls defined. > > Ah, I was using a pretty old version 0.11.07. Now I switched to 0.13.12 > which includes: > > 9038e442b92d - stress-prctl: add Linux 5.14 PR_SCHED_CORE prctl > > To get SCHED_CORE activated in stress-prctl.c, as a quick hack, I had to > add the definitions of PR_SCHED_CORE, PR_SCHED_CORE_GET, etc. to this file. > > Now the issue you described triggers on this machine immediately. > Great! I'll repost the patch with a more accurate commit message then. And if you come up with something different that works for me too. Let me know and I'll test it here. Cheers, Phil -- ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-30 15:52 ` Phil Auld 0 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-30 15:52 UTC (permalink / raw) To: Dietmar Eggemann Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On Wed, Mar 30, 2022 at 05:48:34PM +0200 Dietmar Eggemann wrote: > On 29/03/2022 21:50, Phil Auld wrote: > > On Tue, Mar 29, 2022 at 08:55:08PM +0200 Dietmar Eggemann wrote: > >> On 29/03/2022 17:20, Phil Auld wrote: > >>> On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: > >>>> On 22/03/2022 17:03, Phil Auld wrote: > > [...] > > >>> This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 > >>> chips. > >>> > >>> ARM (CN9980-2200LG4077-Y21-G) > >> I'm using the same processor just with ACPI/PPTT. > >> > > > > Maybe I'm misinformed about these systems having no PPTT... > > > > I'm reclaiming the system. Is there a way I can tell from userspace? > > # cat /sys/firmware/acpi/tables/PPTT > pptt.dat > # iasl -d pptt.dat > # vim pptt.dsl > Thanks, I'll git that a try. I suspect these are the same as yours though and I was just mistaken :) > [...] > > >> so no SMT sched domain. The MPIDR-based topology fallback code in > >> store_cpu_topology() forces `cpuid_topo->thread_id = -1`. > > > > Right. So since I'm getting SMT it must not have package_id == -1. > > In which case you should be able to reproduce it because it must > > be that the call the update_siblings_masks() is required. That > > appears to only be called from store_cpu_topology() which is > > after the scheduler has already setup the core pointers. > > > > The fix could be the same but I should reword the commit message > > since it should effect all SMT arm systems I'd think. > > > > Or maybe the ACPI topology code should call update_sibling_masks(). > >> > >> IMHO this is why on my machine I don't see this issue while running: > >> > >> root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 > >> stress-ng: info: [2388042] dispatching hogs: 256 prctl > >> > >> Is there something I miss in my setup to provoke this issue? > >> > > > > Make sure you have a stress-ng that is new enough and built against > > headers that have the CORE_SCHED prctls defined. > > Ah, I was using a pretty old version 0.11.07. Now I switched to 0.13.12 > which includes: > > 9038e442b92d - stress-prctl: add Linux 5.14 PR_SCHED_CORE prctl > > To get SCHED_CORE activated in stress-prctl.c, as a quick hack, I had to > add the definitions of PR_SCHED_CORE, PR_SCHED_CORE_GET, etc. to this file. > > Now the issue you described triggers on this machine immediately. > Great! I'll repost the patch with a more accurate commit message then. And if you come up with something different that works for me too. Let me know and I'll test it here. Cheers, Phil -- _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling 2022-03-30 15:48 ` Dietmar Eggemann @ 2022-03-30 16:07 ` Phil Auld -1 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-30 16:07 UTC (permalink / raw) To: Dietmar Eggemann Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On Wed, Mar 30, 2022 at 05:48:34PM +0200 Dietmar Eggemann wrote: > On 29/03/2022 21:50, Phil Auld wrote: > > On Tue, Mar 29, 2022 at 08:55:08PM +0200 Dietmar Eggemann wrote: > >> On 29/03/2022 17:20, Phil Auld wrote: > >>> On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: > >>>> On 22/03/2022 17:03, Phil Auld wrote: > > [...] > > >>> This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 > >>> chips. > >>> > >>> ARM (CN9980-2200LG4077-Y21-G) > >> I'm using the same processor just with ACPI/PPTT. > >> > > > > Maybe I'm misinformed about these systems having no PPTT... > > > > I'm reclaiming the system. Is there a way I can tell from userspace? > > # cat /sys/firmware/acpi/tables/PPTT > pptt.dat > # iasl -d pptt.dat > # vim pptt.dsl > I don't have iasl but # strings pptt.dat PPTT ServerCL CAVM So that looks like it has a PPTT entry. Cheers, Phil > [...] > > >> so no SMT sched domain. The MPIDR-based topology fallback code in > >> store_cpu_topology() forces `cpuid_topo->thread_id = -1`. > > > > Right. So since I'm getting SMT it must not have package_id == -1. > > In which case you should be able to reproduce it because it must > > be that the call the update_siblings_masks() is required. That > > appears to only be called from store_cpu_topology() which is > > after the scheduler has already setup the core pointers. > > > > The fix could be the same but I should reword the commit message > > since it should effect all SMT arm systems I'd think. > > > > Or maybe the ACPI topology code should call update_sibling_masks(). > >> > >> IMHO this is why on my machine I don't see this issue while running: > >> > >> root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 > >> stress-ng: info: [2388042] dispatching hogs: 256 prctl > >> > >> Is there something I miss in my setup to provoke this issue? > >> > > > > Make sure you have a stress-ng that is new enough and built against > > headers that have the CORE_SCHED prctls defined. > > Ah, I was using a pretty old version 0.11.07. Now I switched to 0.13.12 > which includes: > > 9038e442b92d - stress-prctl: add Linux 5.14 PR_SCHED_CORE prctl > > To get SCHED_CORE activated in stress-prctl.c, as a quick hack, I had to > add the definitions of PR_SCHED_CORE, PR_SCHED_CORE_GET, etc. to this file. > > Now the issue you described triggers on this machine immediately. > -- ^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH] arch/arm64: Fix topology initialization for core scheduling @ 2022-03-30 16:07 ` Phil Auld 0 siblings, 0 replies; 16+ messages in thread From: Phil Auld @ 2022-03-30 16:07 UTC (permalink / raw) To: Dietmar Eggemann Cc: linux-kernel, Catalin Marinas, Will Deacon, Mark Rutland, Peter Zijlstra, linux-arm-kernel On Wed, Mar 30, 2022 at 05:48:34PM +0200 Dietmar Eggemann wrote: > On 29/03/2022 21:50, Phil Auld wrote: > > On Tue, Mar 29, 2022 at 08:55:08PM +0200 Dietmar Eggemann wrote: > >> On 29/03/2022 17:20, Phil Auld wrote: > >>> On Tue, Mar 29, 2022 at 04:02:22PM +0200 Dietmar Eggemann wrote: > >>>> On 22/03/2022 17:03, Phil Auld wrote: > > [...] > > >>> This instance is an HPE Apollo 70 set to smt-4. I believe it's ThunderX2 > >>> chips. > >>> > >>> ARM (CN9980-2200LG4077-Y21-G) > >> I'm using the same processor just with ACPI/PPTT. > >> > > > > Maybe I'm misinformed about these systems having no PPTT... > > > > I'm reclaiming the system. Is there a way I can tell from userspace? > > # cat /sys/firmware/acpi/tables/PPTT > pptt.dat > # iasl -d pptt.dat > # vim pptt.dsl > I don't have iasl but # strings pptt.dat PPTT ServerCL CAVM So that looks like it has a PPTT entry. Cheers, Phil > [...] > > >> so no SMT sched domain. The MPIDR-based topology fallback code in > >> store_cpu_topology() forces `cpuid_topo->thread_id = -1`. > > > > Right. So since I'm getting SMT it must not have package_id == -1. > > In which case you should be able to reproduce it because it must > > be that the call the update_siblings_masks() is required. That > > appears to only be called from store_cpu_topology() which is > > after the scheduler has already setup the core pointers. > > > > The fix could be the same but I should reword the commit message > > since it should effect all SMT arm systems I'd think. > > > > Or maybe the ACPI topology code should call update_sibling_masks(). > >> > >> IMHO this is why on my machine I don't see this issue while running: > >> > >> root@oss-apollo7007:~# stress-ng --prctl 256 -t 60 > >> stress-ng: info: [2388042] dispatching hogs: 256 prctl > >> > >> Is there something I miss in my setup to provoke this issue? > >> > > > > Make sure you have a stress-ng that is new enough and built against > > headers that have the CORE_SCHED prctls defined. > > Ah, I was using a pretty old version 0.11.07. Now I switched to 0.13.12 > which includes: > > 9038e442b92d - stress-prctl: add Linux 5.14 PR_SCHED_CORE prctl > > To get SCHED_CORE activated in stress-prctl.c, as a quick hack, I had to > add the definitions of PR_SCHED_CORE, PR_SCHED_CORE_GET, etc. to this file. > > Now the issue you described triggers on this machine immediately. > -- _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel ^ permalink raw reply [flat|nested] 16+ messages in thread
end of thread, other threads:[~2022-03-30 16:09 UTC | newest] Thread overview: 16+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2022-03-22 16:03 [PATCH] arch/arm64: Fix topology initialization for core scheduling Phil Auld 2022-03-22 16:03 ` Phil Auld 2022-03-29 14:02 ` Dietmar Eggemann 2022-03-29 14:02 ` Dietmar Eggemann 2022-03-29 15:20 ` Phil Auld 2022-03-29 15:20 ` Phil Auld 2022-03-29 18:55 ` Dietmar Eggemann 2022-03-29 18:55 ` Dietmar Eggemann 2022-03-29 19:50 ` Phil Auld 2022-03-29 19:50 ` Phil Auld 2022-03-30 15:48 ` Dietmar Eggemann 2022-03-30 15:48 ` Dietmar Eggemann 2022-03-30 15:52 ` Phil Auld 2022-03-30 15:52 ` Phil Auld 2022-03-30 16:07 ` Phil Auld 2022-03-30 16:07 ` Phil Auld
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.