s390x: firq: floating interrupt test
cover s390x: firq: floating interrupt test
From patch #2:
"
We had a KVM BUG fixed by kernel commit a3e03bc1368c ("KVM: s390: index kvm->arch.idle_mask by vcpu_idx"), whereby a floating interrupt might get stuck forever because a CPU in the wait state would not get woken up.
The issue can be triggered when CPUs are created in a nonlinear fashion, such that the CPU address ("core-id") and the KVM cpu id don't match.
So let's start with a floating interrupt test that will trigger a floating interrupt (via SCLP) to be delivered to a CPU in the wait state. "
David Hildenbrand (2):
s390x: make smp_cpu_setup() return 0 on success
s390x: firq: floating interrupt test
lib/s390x/smp.c | 1 + s390x/Makefile | 1 + s390x/firq.c | 141 ++++++++++++++++++++++++++++++++++++++++++++ s390x/unittests.cfg | 10 ++++ 4 files changed, 153 insertions(+) create mode 100644 s390x/firq.c
Signature · 1 line
2.31.1
PATCH 1/2 s390x: make smp_cpu_setup() return 0 on success
Properly return "0" on success so callers can check if the setup was successful.
The return value is yet unused, which is why this wasn't noticed so far.
--- lib/s390x/smp.c | 1 + 1 file changed, 1 insertion(+)
lib/s390x/smp.c +1 -0
@@ -212,6 +212,7 @@ int smp_cpu_setup(uint16_t addr, struct psw psw)/* Wait until the cpu has finished setup and started the provided psw */ while (lc->restart_new_psw.addr != psw.addr)mb (); + rc =0 ; out:spin_unlock (&lock);return rc;
Signature · 1 line
2.31.1
Re: [kvm-unit-tests PATCH v1 1/2] s390x: make smp_cpu_setup() return 0 on success
Quoted 16 lines · David Hildenbrand
Properly return "0" on success so callers can check if the setup was successful.
The return value is yet unused, which is why this wasn't noticed so far.
Signed-off-by David Hildenbrand <[email protected]>--- lib/s390x/smp.c | 1 + 1 file changed, 1 insertion(+)lib/s390x/smp.c +0
-0
lib/s390x/smp.c int smp_cpu_setup(uint16_t addr, struct psw psw) · line 218
Quoted 7 lines
lib/s390x/smp.c +1
-0@@ -212,6 +212,7 @@ int smp_cpu_setup(uint16_t addr, struct psw psw)/* Wait until the cpu has finished setup and started the provided psw */ while (lc->restart_new_psw.addr != psw.addr)mb (); + rc =0 ; out:spin_unlock (&lock);return rc;
Re: [kvm-unit-tests PATCH v1 1/2] s390x: make smp_cpu_setup() return 0 on success
In reply to David HildenbrandQuoted 6 lines · On Thu, 2 Dec 2021 10:58:42 +0100 David Hildenbrand <[email protected]> wrote:
Properly return "0" on success so callers can check if the setup was successful.
The return value is yet unused, which is why this wasn't noticed so far.
Signed-off-by David Hildenbrand <[email protected]>
Quoted 16 lines
--- lib/s390x/smp.c | 1 + 1 file changed, 1 insertion(+)lib/s390x/smp.c +1
-0@@ -212,6 +212,7 @@ int smp_cpu_setup(uint16_t addr, struct psw psw)/* Wait until the cpu has finished setup and started the provided psw */ while (lc->restart_new_psw.addr != psw.addr)mb (); + rc =0 ; out:spin_unlock (&lock);return rc;
Re: [kvm-unit-tests PATCH v1 1/2] s390x: make smp_cpu_setup() return 0 on success
In reply to David HildenbrandQuoted 16 lines · David Hildenbrand
Properly return "0" on success so callers can check if the setup was successful.
The return value is yet unused, which is why this wasn't noticed so far.
Signed-off-by David Hildenbrand <[email protected]>--- lib/s390x/smp.c | 1 + 1 file changed, 1 insertion(+)lib/s390x/smp.c +0
-0
lib/s390x/smp.c int smp_cpu_setup(uint16_t addr, struct psw psw) · line 218
Quoted 7 lines
lib/s390x/smp.c +1
-0@@ -212,6 +212,7 @@ int smp_cpu_setup(uint16_t addr, struct psw psw)/* Wait until the cpu has finished setup and started the provided psw */ while (lc->restart_new_psw.addr != psw.addr)mb (); + rc =0 ; out:spin_unlock (&lock);return rc;
oops
PATCH 2/2 s390x: firq: floating interrupt test
In reply to David HildenbrandWe had a KVM BUG fixed by kernel commit a3e03bc1368c ("KVM: s390: index kvm->arch.idle_mask by vcpu_idx"), whereby a floating interrupt might get stuck forever because a CPU in the wait state would not get woken up.
The issue can be triggered when CPUs are created in a nonlinear fashion, such that the CPU address ("core-id") and the KVM cpu id don't match.
So let's start with a floating interrupt test that will trigger a floating interrupt (via SCLP) to be delivered to a CPU in the wait state.
--- s390x/Makefile | 1 + s390x/firq.c | 141 ++++++++++++++++++++++++++++++++++++++++++++ s390x/unittests.cfg | 10 ++++ 3 files changed, 152 insertions(+) create mode 100644 s390x/firq.c
s390x/Makefile +1 -0
@@ -25,6 +25,7 @@ tests += $(TEST_DIR)/uv-host.elf tests += $(TEST_DIR)/edat.elf tests += $(TEST_DIR)/mvpg-sie.elf tests += $(TEST_DIR)/spec_ex-sie.elf +tests += $(TEST_DIR)/firq.elf tests_binary = $(patsubst %.elf,%.bin,$(tests)) ifneq ($(HOST_KEY_DOCUMENT),)
s390x/firq.c new file +141 -0
141 lines not on this page · raw message
s390x/unittests.cfg +10 -0
@@ -112,3 +112,13 @@ file = mvpg-sie.elf [spec_ex-sie] file = spec_ex-sie.elf + +[firq-linear-cpu-ids] +file = firq.elf +timeout = 20 +extra_params = -smp 1,maxcpus=3 -cpu qemu -device qemu-s390x-cpu,core-id=1 -device qemu-s390x-cpu,core-id=2 + +[firq-nonlinear-cpu-ids] +file = firq.elf +timeout = 20 +extra_params = -smp 1,maxcpus=3 -cpu qemu -device qemu-s390x-cpu,core-id=2 -device qemu-s390x-cpu,core-id=1
Signature · 1 line
2.31.1
Quoted 66 lines · On Thu, 2 Dec 2021 10:58:43 +0100 David Hildenbrand <[email protected]> wrote:
We had a KVM BUG fixed by kernel commit
a3e03bc1368c("KVM: s390: indexkvm->arch.idle_maskbyvcpu_idx"), whereby a floating interrupt might get stuck forever because a CPU in the wait state would not get woken up.The issue can be triggered when CPUs are created in a nonlinear fashion, such that the CPU address ("core-id") and the KVM cpu id don't match.
So let's start with a floating interrupt test that will trigger a floating interrupt (via SCLP) to be delivered to a CPU in the wait state.
Signed-off-by David Hildenbrand <[email protected]>--- s390x/Makefile | 1 + s390x/firq.c | 141 ++++++++++++++++++++++++++++++++++++++++++++ s390x/unittests.cfg | 10 ++++ 3 files changed, 152 insertions(+) create mode 100644 s390x/firq.cs390x/Makefile +1
-0@@ -25,6 +25,7 @@ tests += $(TEST_DIR)/uv-host.elf tests += $(TEST_DIR)/edat.elf tests += $(TEST_DIR)/mvpg-sie.elf tests += $(TEST_DIR)/spec_ex-sie.elf +tests += $(TEST_DIR)/firq.elf tests_binary = $(patsubst %.elf,%.bin,$(tests)) ifneq ($(HOST_KEY_DOCUMENT),)s390x/firq.c new file +30
-0@@ -0,0 +1,141 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* +* Floating interrupt tests. +* +* Copyright 2021 Red Hat Inc +* +* Authors: +* David Hildenbrand <[email protected]> +*/ +#include <libcflat.h> +#include <asm/asm-offsets.h> +#include <asm/interrupt.h> +#include <asm/page.h> +#include <asm-generic/barrier.h> + +#include <sclp.h> +#include <smp.h> +#include <alloc_page.h> + +static int testflag =0 ; + +static void wait_for_flag (void ) +{ +while (!testflag) +mb (); +} + +static void set_flag (int val) +{ +mb ();
s390x/firq.c line 40
Quoted 10 lines
s390x/firq.c new file +10
-0+ testflag = val; +mb (); +} + +static void wait_for_sclp_int (void ) +{ +/* Enable SCLP interrupts on this CPU only. */ +ctl_set_bit (0 ,CTL0_SERVICE_SIGNAL ); + +set_flag (1 );
why not just WRITE_ONCE/READ_ONCE?
Quoted 33 lines
s390x/firq.c new file +33
-0+ +/* Enable external interrupts and go to the wait state. */ +wait_for_interrupt (PSW_MASK_EXT ); +} + +/* +* Some KVM versions might mix CPUs when looking for a floating IRQ target, +* accidentially detecting a stopped CPU as waiting and resulting in the actually +* waiting CPU not getting woken up for the interrupt. +*/ +static void test_wait_state_delivery (void ) +{ +struct psw psw; +SCCBHeader *h; +int ret; + +report_prefix_push ("wait state delivery" ); + +if (smp_query_num_cpus () <3 ) { +report_skip ("need at least 3 CPUs for this test" ); + goto out; + } + +if (stap ()) { +report_skip ("need to start on CPU #0" ); + goto out; + } + +/* +* We want CPU #2 to be stopped. This should be the case at this +* point, however, we want to sense if it even exists as well. +*/ + ret =smp_cpu_stop (2 );
s390x/firq.c line 83
Quoted 10 lines
s390x/firq.c new file +10
-0+if (ret) { +report_skip ("CPU #2 not found" ); + goto out; + } + +/* +* We're going to perform an SCLP service call but expect +* the interrupt on CPU #1 while it is in the wait state. +*/ +sclp_mark_busy ();
this means that now no SCLP command can happen as long as the flag stays set....
s390x/firq.c line 91
Quoted 8 lines
s390x/firq.c new file +8
-0+set_flag (0 ); + +/* Start CPU #1 and let it wait for the interrupt. */ + psw.mask =extract_psw_mask (); + psw.addr = (unsigned long )wait_for_sclp_int; + ret =smp_cpu_setup (1 , psw); +if (ret) { +report_skip ("cpu #1 not found" );
...which means that this will hang, and so will all the other report* functions. maybe you should manually unset the flag before calling the various report* functions.
Quoted 8 lines
s390x/firq.c new file +8
-0+ goto out; + } + +/* Wait until the CPU #1 at least enabled SCLP interrupts. */ +wait_for_flag (); + +/* +* We'd have to jump trough some hoops to sense e.g., via SIGP
s390x/firq.c line 109
Quoted 10 lines
s390x/firq.c new file +10
-0+* CONDITIONAL EMERGENCY SIGNAL if CPU #1 is already in the +* wait state. +* +* Although not completely reliable, use SIGP SENSE RUNNING STATUS +* until not reported as running -- after all, our SCLP processing +* will take some time as well and make races very rare. +*/ +while (smp_sense_running_status (1 )); + + h =alloc_page ();
do you really need to dynamically allocate one page?
is there a reason for not using a simple static buffer? (which you can have aligned and statically initialized)
s390x/firq.c line 110
Quoted 1 line
s390x/firq.c new file +1
-0+memset (h,0 ,sizeof (*h));
otherwise, if you really want to allocate the memory, get rid of the memset; the allocator always returns zeroed memory (unless you explicitly ask not to by using flags)
s390x/firq.c line 119
Quoted 9 lines
s390x/firq.c new file +9
-0+ h->length =4096 ; + ret =servc (SCLP_CMDW_READ_CPU_INFO ,__pa (h)); +if (ret) { +report_fail ("SCLP_CMDW_READ_CPU_INFO failed" ); + goto out_destroy; + } + +/* +* Wait until the interrupt gets delivered on CPU #1, marking the
why do you expect the interrupt to be delivered on CPU1? could it not be delivered on CPU0?
s390x/firq.c line 122
Quoted 3 lines
s390x/firq.c new file +3
-0+* SCLP requests as done. +*/ +sclp_wait_busy ();
this is logically not wrong (and should stay, because it makes clear what you are trying to do), but strictly speaking it's not needed since the report below will hang as long as the SCLP busy flag is set.
Quoted 37 lines
s390x/firq.c new file +19
-0+ +report (true,"firq delivered" ); + +out_destroy: +smp_cpu_destroy (1 ); +free_page (h); +out: +report_prefix_pop (); +} + +int main (void ) +{ +report_prefix_push ("firq" ); + +test_wait_state_delivery (); + +report_prefix_pop (); +return report_summary (); +}s390x/unittests.cfg +10
-0@@ -112,3 +112,13 @@ file = mvpg-sie.elf [spec_ex-sie] file = spec_ex-sie.elf + +[firq-linear-cpu-ids] +file = firq.elf +timeout = 20 +extra_params = -smp 1,maxcpus=3 -cpu qemu -device qemu-s390x-cpu,core-id=1 -device qemu-s390x-cpu,core-id=2 + +[firq-nonlinear-cpu-ids] +file = firq.elf +timeout = 20 +extra_params = -smp 1,maxcpus=3 -cpu qemu -device qemu-s390x-cpu,core-id=2 -device qemu-s390x-cpu,core-id=1
Quoted 8 lines
s390x/firq.c line 40
Quoted 6 lines
s390x/firq.c new file +6
-0+static void wait_for_sclp_int (void ) +{ +/* Enable SCLP interrupts on this CPU only. */ +ctl_set_bit (0 ,CTL0_SERVICE_SIGNAL ); + +set_flag (1 );why not just WRITE_ONCE/READ_ONCE?
Because I shamelessly copied that from s390x/smp.c ;)
Quoted 4 lines
Quoted 4 lines
s390x/firq.c new file +4
-0+set_flag (0 ); + +/* Start CPU #1 and let it wait for the interrupt. */ + psw.mask =extract_psw_mask ();
Quoted 8 lines
s390x/firq.c line 91
Quoted 4 lines
s390x/firq.c new file +4
-0+ psw.addr = (unsigned long )wait_for_sclp_int; + ret =smp_cpu_setup (1 , psw); +if (ret) { +report_skip ("cpu #1 not found" );...which means that this will hang, and so will all the other report* functions. maybe you should manually unset the flag before calling the various report* functions.
Good point, thanks!
Quoted 15 lines
Quoted 15 lines
s390x/firq.c new file +14
-0+ goto out; + } + +/* Wait until the CPU #1 at least enabled SCLP interrupts. */ +wait_for_flag (); + +/* +* We'd have to jump trough some hoops to sense e.g., via SIGP +* CONDITIONAL EMERGENCY SIGNAL if CPU #1 is already in the +* wait state. +* +* Although not completely reliable, use SIGP SENSE RUNNING STATUS +* until not reported as running -- after all, our SCLP processing +* will take some time as well and make races very rare.
Quoted 8 lines
s390x/firq.c line 109
Quoted 4 lines
s390x/firq.c new file +4
-0+*/ +while (smp_sense_running_status (1 )); + + h =alloc_page ();do you really need to dynamically allocate one page?
is there a reason for not using a simple static buffer? (which you can have aligned and statically initialized)
I don't really have a strong opinion. I do prefer dynamic alloctions, though, if there isn't a good reason not to use them. No need to mess with page alignments manually.
Quoted 6 lines
s390x/firq.c line 110
Quoted 2 lines
s390x/firq.c new file +1
-0+memset (h,0 ,sizeof (*h));otherwise, if you really want to allocate the memory, get rid of the memset; the allocator always returns zeroed memory (unless you explicitly ask not to by using flags)
Right. "special" FLAG_DONTZERO in that semantics in that allocator.
Quoted 13 lines
s390x/firq.c line 119
Quoted 10 lines
s390x/firq.c new file +9
-0+ h->length =4096 ; + ret =servc (SCLP_CMDW_READ_CPU_INFO ,__pa (h)); +if (ret) { +report_fail ("SCLP_CMDW_READ_CPU_INFO failed" ); + goto out_destroy; + } + +/* +* Wait until the interrupt gets delivered on CPU #1, marking the why do you expect the interrupt to be delivered on CPU1? could it not be delivered on CPU0?
We don't enable SCLP interrupts + external interrupts on CPU #0 because we'll only call sclp_setup_int() on CPU #1.
Quoted 8 lines
s390x/firq.c line 122
Quoted 4 lines
s390x/firq.c new file +3
-0+* SCLP requests as done. +*/ +sclp_wait_busy ();this is logically not wrong (and should stay, because it makes clear what you are trying to do), but strictly speaking it's not needed since the report below will hang as long as the SCLP busy flag is set.
Right. But it's really clearer to just have this in the code.
Signature · 3 lines
Thanks, David / dhildenb
Quoted 10 lines · David Hildenbrand
Quoted 8 lines
s390x/firq.c line 40
Quoted 6 lines
s390x/firq.c new file +6
-0+static void wait_for_sclp_int (void ) +{ +/* Enable SCLP interrupts on this CPU only. */ +ctl_set_bit (0 ,CTL0_SERVICE_SIGNAL ); + +set_flag (1 );why not just WRITE_ONCE/READ_ONCE?
Because I shamelessly copied that from
s390x/smp.c;)
Oh, and also because WRITE_ONCE/READ_ONCE are semantically the wrong thing to use:
"Prevent the compiler from merging or refetching reads or writes. The compiler is also forbidden from reordering successive instances of READ_ONCE and WRITE_ONCE"
We need memory barriers to prevent reordering with surrounding code.
Signature · 3 lines
Thanks, David / dhildenb
Quoted 27 lines · On Thu, 2 Dec 2021 12:13:08 +0100 David Hildenbrand <[email protected]> wrote:
Quoted 8 lines
s390x/firq.c line 40
Quoted 6 lines
s390x/firq.c new file +6
-0+static void wait_for_sclp_int (void ) +{ +/* Enable SCLP interrupts on this CPU only. */ +ctl_set_bit (0 ,CTL0_SERVICE_SIGNAL ); + +set_flag (1 );why not just WRITE_ONCE/READ_ONCE?
Because I shamelessly copied that from
s390x/smp.c;)Quoted 4 lines
Quoted 4 lines
s390x/firq.c new file +4
-0+set_flag (0 ); + +/* Start CPU #1 and let it wait for the interrupt. */ + psw.mask =extract_psw_mask ();Quoted 8 lines
s390x/firq.c line 91
Quoted 4 lines
s390x/firq.c new file +4
-0+ psw.addr = (unsigned long )wait_for_sclp_int; + ret =smp_cpu_setup (1 , psw); +if (ret) { +report_skip ("cpu #1 not found" );...which means that this will hang, and so will all the other report* functions. maybe you should manually unset the flag before calling the various report* functions.
Good point, thanks!
Quoted 1 line
Quoted 16 lines
Quoted 8 lines
Quoted 8 lines
s390x/firq.c new file +8
-0+ goto out; + } + +/* Wait until the CPU #1 at least enabled SCLP interrupts. */ +wait_for_flag (); + +/* +* We'd have to jump trough some hoops to sense e.g., via SIGP Quoted 8 lines
s390x/firq.c line 107
Quoted 8 lines
s390x/firq.c new file +8
-0+* CONDITIONAL EMERGENCY SIGNAL if CPU #1 is already in the +* wait state. +* +* Although not completely reliable, use SIGP SENSE RUNNING STATUS +* until not reported as running -- after all, our SCLP processing +* will take some time as well and make races very rare. +*/ +while (smp_sense_running_status (1 ));
if you wait here for CPU1 to be in wait state, then why did you need to wait until it has set the flag earlier? can't you just wait here and not use the whole wait_for_flag logic? smp_cpu_setup only returns after the new CPU has started running.
I guess this was also inspired by smp.c :)
Quoted 10 lines
Quoted 6 lines
Review
Quoted 2 lines
(quoted diff) +2
-0+ + h = alloc_page();do you really need to dynamically allocate one page?
is there a reason for not using a simple static buffer? (which you can have aligned and statically initialized)I don't really have a strong opinion. I do prefer dynamic alloctions, though, if there isn't a good reason not to use them. No need to mess with page alignments manually.
fair enough, I also do not have a strong opinion
Quoted 23 lines
Quoted 6 lines
s390x/firq.c line 110
Quoted 2 lines
s390x/firq.c new file +1
-0+memset (h,0 ,sizeof (*h));otherwise, if you really want to allocate the memory, get rid of the memset; the allocator always returns zeroed memory (unless you explicitly ask not to by using flags)
Right. "special"
FLAG_DONTZEROin that semantics in that allocator.Quoted 13 lines
s390x/firq.c line 119
Quoted 10 lines
s390x/firq.c new file +9
-0+ h->length =4096 ; + ret =servc (SCLP_CMDW_READ_CPU_INFO ,__pa (h)); +if (ret) { +report_fail ("SCLP_CMDW_READ_CPU_INFO failed" ); + goto out_destroy; + } + +/* +* Wait until the interrupt gets delivered on CPU #1, marking the why do you expect the interrupt to be delivered on CPU1? could it not be delivered on CPU0?
Quoted 2 lines
We don't enable SCLP interrupts + external interrupts on CPU #0 because we'll only call
sclp_setup_int()on CPU #1.
oh right, I had missed that
Quoted 11 lines
Quoted 9 lines
s390x/firq.c line 122
Quoted 5 lines
s390x/firq.c new file +3
-0+* SCLP requests as done. +*/ +sclp_wait_busy ();this is logically not wrong (and should stay, because it makes clear what you are trying to do), but strictly speaking it's not needed since the report below will hang as long as the SCLP busy flag is set.
Right. But it's really clearer to just have this in the code.
yep, which is why I said that this should stay :)
Quoted 2 lines
Quoted 43 lines · Claudio Imbrenda
Quoted 40 lines · On Thu, 2 Dec 2021 12:13:08 +0100 David Hildenbrand <[email protected]> wrote:
Quoted 8 lines
s390x/firq.c line 40
Quoted 6 lines
s390x/firq.c new file +6
-0+static void wait_for_sclp_int (void ) +{ +/* Enable SCLP interrupts on this CPU only. */ +ctl_set_bit (0 ,CTL0_SERVICE_SIGNAL ); + +set_flag (1 );why not just WRITE_ONCE/READ_ONCE?
Because I shamelessly copied that from
s390x/smp.c;)Quoted 4 lines
Quoted 4 lines
s390x/firq.c new file +4
-0+set_flag (0 ); + +/* Start CPU #1 and let it wait for the interrupt. */ + psw.mask =extract_psw_mask ();Quoted 8 lines
s390x/firq.c line 91
Quoted 4 lines
s390x/firq.c new file +4
-0+ psw.addr = (unsigned long )wait_for_sclp_int; + ret =smp_cpu_setup (1 , psw); +if (ret) { +report_skip ("cpu #1 not found" );...which means that this will hang, and so will all the other report* functions. maybe you should manually unset the flag before calling the various report* functions.
Good point, thanks!
Quoted 14 lines
Quoted 14 lines
s390x/firq.c new file +13
-0+ goto out; + } + +/* Wait until the CPU #1 at least enabled SCLP interrupts. */ +wait_for_flag (); + +/* +* We'd have to jump trough some hoops to sense e.g., via SIGP +* CONDITIONAL EMERGENCY SIGNAL if CPU #1 is already in the +* wait state. +* +* Although not completely reliable, use SIGP SENSE RUNNING STATUS +* until not reported as running -- after all, our SCLP processing
Quoted 8 lines
Quoted 3 lines
Quoted 3 lines
s390x/firq.c line 107
Quoted 3 lines
s390x/firq.c new file +3
-0+* will take some time as well and make races very rare. +*/ +while (smp_sense_running_status (1 ));if you wait here for CPU1 to be in wait state, then why did you need to wait until it has set the flag earlier? can't you just wait here and not use the whole
wait_for_flaglogic?smp_cpu_setuponly returns after the new CPU has started running.
I use the flag right now as a mechanism to make the race window as short as possible. But you're right, we might not need the flag logic as knowing that we processed the setup part and will jump/jumped to the target code might be good enough.
Thanks!
Signature · 3 lines
Thanks, David / dhildenb