| Message ID | 1543404287-3946-1-git-send-email-hofrat@osadl.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | perf: arm_spe: handle devm_kasprintf() failure | expand |
On Wed, Nov 28, 2018 at 12:24:47PM +0100, Nicholas Mc Guire wrote: > devm_kasprintf() may return NULL on failure of internal allocation > thus the assignment to name is not safe if unchecked. if NULL > is passed in for name then perf_pmu_register() would not fail > but rather silently jump to skip_type which is not the intent > here. As perf_pmu_register() may also return -ENOMEM returning > -ENOMEM in the (unlikely) failure case of devm_kasprintf() should > be fine here as well. > > Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org> > Fixes: d5d9696b0380 ("drivers/perf: Add support for ARMv8.2 Statistical Profiling Extension") > --- > > Problem located with an experimental coccinelle script > > The dev_err() might seem a bit much for this unlikely error but as > perf_pmu_register() may also return -ENOMEM it would be hard to figure > out what went wrong without this message. > > Patch was compile tested with: defconfig (ARCH=arm64) + > ARM_SPE_PMU=y > > Patch is against 4.20-rc3 (localversion-next is next-20181128) > > drivers/perf/arm_spe_pmu.c | 5 +++++ > 1 file changed, 5 insertions(+) > > diff --git a/drivers/perf/arm_spe_pmu.c b/drivers/perf/arm_spe_pmu.c > index 54ec278..f1ea00c 100644 > --- a/drivers/perf/arm_spe_pmu.c > +++ b/drivers/perf/arm_spe_pmu.c > @@ -927,6 +927,11 @@ static int arm_spe_pmu_perf_init(struct arm_spe_pmu *spe_pmu) > > idx = atomic_inc_return(&pmu_idx); > name = devm_kasprintf(dev, GFP_KERNEL, "%s_%d", PMUNAME, idx); > + if (!name) { > + dev_err(dev, "Allocation of name failed\n"); > + return -ENOMEM; > + } In the other dev_err() messages, we use the form "failed to ...", e.g. "failed to get IRQ (%d)\n" "failed to get PPI partition (%d)\n "failed to allocate spe_pmu\n" ... so for consistency could you please change the string to: "failed to allocate spe_pmu name\n" With that, feel free to add: Acked-by: Mark Rutland <mark.rutland@arm.com> Mark.
On Thu, Nov 29, 2018 at 12:14:32PM +0000, Mark Rutland wrote: > On Wed, Nov 28, 2018 at 12:24:47PM +0100, Nicholas Mc Guire wrote: > > diff --git a/drivers/perf/arm_spe_pmu.c b/drivers/perf/arm_spe_pmu.c > > index 54ec278..f1ea00c 100644 > > --- a/drivers/perf/arm_spe_pmu.c > > +++ b/drivers/perf/arm_spe_pmu.c > > @@ -927,6 +927,11 @@ static int arm_spe_pmu_perf_init(struct arm_spe_pmu *spe_pmu) > > > > idx = atomic_inc_return(&pmu_idx); > > name = devm_kasprintf(dev, GFP_KERNEL, "%s_%d", PMUNAME, idx); > > + if (!name) { > > + dev_err(dev, "Allocation of name failed\n"); > > + return -ENOMEM; > > + } > > In the other dev_err() messages, we use the form "failed to ...", e.g. > > "failed to get IRQ (%d)\n" > "failed to get PPI partition (%d)\n > "failed to allocate spe_pmu\n" > > ... so for consistency could you please change the string to: > > "failed to allocate spe_pmu name\n" > > With that, feel free to add: > > Acked-by: Mark Rutland <mark.rutland@arm.com> Thanks, I've queued this for 4.21 with a tweaked error message. Will
diff --git a/drivers/perf/arm_spe_pmu.c b/drivers/perf/arm_spe_pmu.c index 54ec278..f1ea00c 100644 --- a/drivers/perf/arm_spe_pmu.c +++ b/drivers/perf/arm_spe_pmu.c @@ -927,6 +927,11 @@ static int arm_spe_pmu_perf_init(struct arm_spe_pmu *spe_pmu) idx = atomic_inc_return(&pmu_idx); name = devm_kasprintf(dev, GFP_KERNEL, "%s_%d", PMUNAME, idx); + if (!name) { + dev_err(dev, "Allocation of name failed\n"); + return -ENOMEM; + } + return perf_pmu_register(&spe_pmu->pmu, name, -1); }
devm_kasprintf() may return NULL on failure of internal allocation thus the assignment to name is not safe if unchecked. if NULL is passed in for name then perf_pmu_register() would not fail but rather silently jump to skip_type which is not the intent here. As perf_pmu_register() may also return -ENOMEM returning -ENOMEM in the (unlikely) failure case of devm_kasprintf() should be fine here as well. Signed-off-by: Nicholas Mc Guire <hofrat@osadl.org> Fixes: d5d9696b0380 ("drivers/perf: Add support for ARMv8.2 Statistical Profiling Extension") --- Problem located with an experimental coccinelle script The dev_err() might seem a bit much for this unlikely error but as perf_pmu_register() may also return -ENOMEM it would be hard to figure out what went wrong without this message. Patch was compile tested with: defconfig (ARCH=arm64) + ARM_SPE_PMU=y Patch is against 4.20-rc3 (localversion-next is next-20181128) drivers/perf/arm_spe_pmu.c | 5 +++++ 1 file changed, 5 insertions(+)