From mboxrd@z Thu Jan 1 00:00:00 1970 From: ming.lei@redhat.com (Ming Lei) Date: Wed, 14 Aug 2019 16:58:46 +0800 Subject: [PATCH V3 1/3] genirq/affinity: Enhance warning check In-Reply-To: References: <20190813081447.1396-1-ming.lei@redhat.com> <20190813081447.1396-2-ming.lei@redhat.com> Message-ID: <20190814085845.GA21802@ming.t460p> On Tue, Aug 13, 2019@07:31:39PM +0000, Derrick, Jonathan wrote: > Hi Ming, > > On Tue, 2019-08-13@16:14 +0800, Ming Lei wrote: > > The two-stage spread is done on same irq vectors, and we just need that > > either one stage covers all vector, not two stage work together to cover > > all vectors. > > > > So enhance the warning check to make sure all vectors are spread. > > > > Cc: Christoph Hellwig > > Cc: Keith Busch > > Cc: linux-nvme at lists.infradead.org, > > Cc: Jon Derrick > > Cc: Jens Axboe > > Fixes: 6da4b3ab9a6 ("genirq/affinity: Add support for allocating interrupt sets") > > Signed-off-by: Ming Lei > > --- > > kernel/irq/affinity.c | 3 +-- > > 1 file changed, 1 insertion(+), 2 deletions(-) > > > > diff --git a/kernel/irq/affinity.c b/kernel/irq/affinity.c > > index 6fef48033f96..265b3076f16b 100644 > > --- a/kernel/irq/affinity.c > > +++ b/kernel/irq/affinity.c > > @@ -215,8 +215,7 @@ static int irq_build_affinity_masks(unsigned int startvec, unsigned int numvecs, > > npresmsk, nmsk, masks); > > put_online_cpus(); > > > > - if (nr_present < numvecs) > > - WARN_ON(nr_present + nr_others < numvecs); > > + WARN_ON(max(nr_present, nr_others) < numvecs); > > I think the patch description assumes the first condition > "The two-stage spread is done on same irq vectors" > > /* > * Spread on non present CPUs starting from the next vector to be > * handled. If the spreading of present CPUs already exhausted the > * vector space, assign the non present CPUs to the already spread > * out vectors. > */ > if (nr_present >= numvecs) > curvec = firstvec; > > But doesn't following condition imply nr_others spread is potentionally > different vector set? > > else > curvec = firstvec + nr_present; > Most times, __irq_build_affinity_masks() returns numvecs. However, each stage may expose less CPU number than num_vecs, then less vectors than 'numvecs' can be spread. So this patch is actually wrong. Thank, Ming