On 3/3/22 22:04, Eric Farman wrote:
A SIGP SENSE is used to determine if a CPU is stopped or operating,
and thus has a vested interest in ensuring it received a CC0 or CC1,
instead of a CC2 (BUSY). But, any order could receive a CC2 response,
and is probably ill-equipped to respond to it.
sigp sense running status doesn't return a cc2, only sigp sense does afaik.
Looking at the KVM implementation tells me that it's not doing more than
looking at the R bit in the sblk.
In practice, the order is likely to only encounter this when racing
with a SIGP STOP (AND STORE STATUS) or SIGP RESTART order, which are
unlikely. But, since it's not impossible, let's convert the library
calls that issue a SIGP to loop on CC2 so the callers do not need
to react to that possible outcome.
Signed-off-by: Eric Farman <farman@xxxxxxxxxxxxx>
---
lib/s390x/smp.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/s390x/smp.c b/lib/s390x/smp.c
index 85b046a5..2e476264 100644
--- a/lib/s390x/smp.c
+++ b/lib/s390x/smp.c
@@ -85,7 +85,7 @@ bool smp_cpu_stopped(uint16_t idx)
bool smp_sense_running_status(uint16_t idx)
{
- if (smp_sigp(idx, SIGP_SENSE_RUNNING, 0, NULL) != SIGP_CC_STATUS_STORED)
+ if (smp_sigp_retry(idx, SIGP_SENSE_RUNNING, 0, NULL) != SIGP_CC_STATUS_STORED)
return true;
/* Status stored condition code is equivalent to cpu not running. */
return false;
@@ -169,7 +169,7 @@ static int smp_cpu_restart_nolock(uint16_t idx, struct psw *psw)
* running after the restart.
*/
smp_cpu_stop_nolock(idx, false);
- rc = smp_sigp(idx, SIGP_RESTART, 0, NULL);
+ rc = smp_sigp_retry(idx, SIGP_RESTART, 0, NULL);
if (rc)
return rc;
/*
