On Mon, Mar 6, 2023 at 3:20 AM Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
>
> Your final commit 596ff4a09b898179 ("cpumask: re-introduce
> constant-sized cpumask optimizations") in v6.3-rc1 introduced a
> regression. During Debian userspace startup, the kernel crashes with:
I'm pretty sure the attached patch should fix it. If you can confirm,
that would be lovely.
The only relevant part is actually just the oneliner to
drivers/char/random.c - the others are fixing the exact same problem
elsewhere, but that's not the case you're actually hitting.
> Presumably using small_cpumask_bits instead of nr_cpu_ids accesses
> some uninitialized array members?
No, it's actually the other way around - some drivers end up using
that "nr_cpumask_bits" in invalid ways, and the small_cpumask_bits
optimization then made that just _very_ obvious.
The bug was pre-existing, it's just that you couldn't trigger it before.
> A similar kernel on an arm64 system that does have 8 CPU cores works fine.
> On an arm64 system with 2 CPU cores, it crashes in a similar way.
Yeah, it's rather machine-specific, and that's why I never saw it in
my local testing either.
So on my machine I always either filled up the cpumask entirely
(because I did my testing on my beefy desktop), or NR_CPUS was so
large that the problem case never happened in the first place because
nr_cpumask_bits was the same as small_cpumask_bits.
I thought I had tested the interesting cases, but in this case, the
case that fails is actually the really trivial case. It's just that my
big machine would never trigger it, because it has so many cores.
Linus
arch/powerpc/xmon/xmon.c | 2 +-
drivers/char/random.c | 2 +-
drivers/net/wireguard/queueing.h | 2 +-
drivers/scsi/lpfc/lpfc_init.c | 14 +++++++-------
4 files changed, 10 insertions(+), 10 deletions(-)
diff --git a/arch/powerpc/xmon/xmon.c b/arch/powerpc/xmon/xmon.c
index 73c620c2a3a1..e753a6bd4888 100644
--- a/arch/powerpc/xmon/xmon.c
+++ b/arch/powerpc/xmon/xmon.c
@@ -1275,7 +1275,7 @@ static int xmon_batch_next_cpu(void)
while (!cpumask_empty(&xmon_batch_cpus)) {
cpu = cpumask_next_wrap(smp_processor_id(), &xmon_batch_cpus,
xmon_batch_start_cpu, true);
- if (cpu == nr_cpumask_bits)
+ if (cpu >= nr_cpu_ids)
break;
if (xmon_batch_start_cpu == -1)
xmon_batch_start_cpu = cpu;
diff --git a/drivers/char/random.c b/drivers/char/random.c
index ce3ccd172cc8..253f2ddb8913 100644
--- a/drivers/char/random.c
+++ b/drivers/char/random.c
@@ -1311,7 +1311,7 @@ static void __cold try_to_generate_entropy(void)
/* Basic CPU round-robin, which avoids the current CPU. */
do {
cpu = cpumask_next(cpu, &timer_cpus);
- if (cpu == nr_cpumask_bits)
+ if (cpu >= nr_cpu_ids)
cpu = cpumask_first(&timer_cpus);
} while (cpu == smp_processor_id() && num_cpus > 1);
diff --git a/drivers/net/wireguard/queueing.h b/drivers/net/wireguard/queueing.h
index 583adb37ee1e..125284b346a7 100644
--- a/drivers/net/wireguard/queueing.h
+++ b/drivers/net/wireguard/queueing.h
@@ -106,7 +106,7 @@ static inline int wg_cpumask_choose_online(int *stored_cpu, unsigned int id)
{
unsigned int cpu = *stored_cpu, cpu_index, i;
- if (unlikely(cpu == nr_cpumask_bits ||
+ if (unlikely(cpu >= nr_cpu_ids ||
!cpumask_test_cpu(cpu, cpu_online_mask))) {
cpu_index = id % cpumask_weight(cpu_online_mask);
cpu = cpumask_first(cpu_online_mask);
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 61958a24a43d..73b544bfbb2e 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -12563,7 +12563,7 @@ lpfc_cpu_affinity_check(struct lpfc_hba *phba, int vectors)
goto found_same;
new_cpu = cpumask_next(
new_cpu, cpu_present_mask);
- if (new_cpu == nr_cpumask_bits)
+ if (new_cpu >= nr_cpu_ids)
new_cpu = first_cpu;
}
/* At this point, we leave the CPU as unassigned */
@@ -12577,7 +12577,7 @@ lpfc_cpu_affinity_check(struct lpfc_hba *phba, int vectors)
* selecting the same IRQ.
*/
start_cpu = cpumask_next(new_cpu, cpu_present_mask);
- if (start_cpu == nr_cpumask_bits)
+ if (start_cpu >= nr_cpu_ids)
start_cpu = first_cpu;
lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
@@ -12613,7 +12613,7 @@ lpfc_cpu_affinity_check(struct lpfc_hba *phba, int vectors)
goto found_any;
new_cpu = cpumask_next(
new_cpu, cpu_present_mask);
- if (new_cpu == nr_cpumask_bits)
+ if (new_cpu >= nr_cpu_ids)
new_cpu = first_cpu;
}
/* We should never leave an entry unassigned */
@@ -12631,7 +12631,7 @@ lpfc_cpu_affinity_check(struct lpfc_hba *phba, int vectors)
* selecting the same IRQ.
*/
start_cpu = cpumask_next(new_cpu, cpu_present_mask);
- if (start_cpu == nr_cpumask_bits)
+ if (start_cpu >= nr_cpu_ids)
start_cpu = first_cpu;
lpfc_printf_log(phba, KERN_INFO, LOG_INIT,
@@ -12704,7 +12704,7 @@ lpfc_cpu_affinity_check(struct lpfc_hba *phba, int vectors)
goto found_hdwq;
}
new_cpu = cpumask_next(new_cpu, cpu_present_mask);
- if (new_cpu == nr_cpumask_bits)
+ if (new_cpu >= nr_cpu_ids)
new_cpu = first_cpu;
}
@@ -12719,7 +12719,7 @@ lpfc_cpu_affinity_check(struct lpfc_hba *phba, int vectors)
goto found_hdwq;
new_cpu = cpumask_next(new_cpu, cpu_present_mask);
- if (new_cpu == nr_cpumask_bits)
+ if (new_cpu >= nr_cpu_ids)
new_cpu = first_cpu;
}
@@ -12730,7 +12730,7 @@ lpfc_cpu_affinity_check(struct lpfc_hba *phba, int vectors)
found_hdwq:
/* We found an available entry, copy the IRQ info */
start_cpu = cpumask_next(new_cpu, cpu_present_mask);
- if (start_cpu == nr_cpumask_bits)
+ if (start_cpu >= nr_cpu_ids)
start_cpu = first_cpu;
cpup->hdwq = new_cpup->hdwq;
logit:
