[RFC PATCH 3/3] nvme-pci: defer completions when rescheduling is needed

[Naman Jain]

On 10/9/2026 9:00 PM, Keith Busch wrote:
> On Fri, Oct 09, 2026 at 05:05:56AM +0000, Naman Jain wrote:
>>   static irqreturn_t nvme_irq(int irq, void *data)
>>   {
>>   	struct nvme_queue *nvmeq = data;
>>   	DEFINE_IO_COMP_BATCH(iob);
>>   
>> +	if (unlikely(test_and_clear_bit(NVMEQ_IRQ_POLL_HANDOFF,
>> +					&nvmeq->flags))) {
>> +		if (!nvme_cqe_pending(nvmeq))
>> +			return IRQ_HANDLED;
>> +	}
>> +	if (unlikely(test_bit(NVMEQ_IRQ_POLL_PAUSED, &nvmeq->flags) &&
>> +		     (!test_bit(NVMEQ_IRQ_POLL_INIT, &nvmeq->flags) ||
>> +		      test_bit(NVMEQ_IRQ_POLL_MASKED, &nvmeq->flags) ||
>> +		      READ_ONCE(to_pci_dev(nvmeq->dev->dev)->error_state) !=
>> +				pci_channel_io_normal)))
>> +		return IRQ_HANDLED;
>> +
>> +	if (unlikely(in_hardirq() &&
>> +		     test_bit(NVMEQ_IRQ_POLL_INIT, &nvmeq->flags) &&
>> +		     !test_bit(NVMEQ_IRQ_POLL_PAUSED, &nvmeq->flags) &&
>> +		     need_resched() && nvme_cqe_pending(nvmeq) &&
>> +		     !test_and_set_bit_lock(NVMEQ_IRQ_POLL, &nvmeq->flags))) {
>> +		set_bit(NVMEQ_IRQ_POLL_MASKED, &nvmeq->flags);
>> +		disable_irq_nosync(irq);
>> +		irq_poll_sched(&nvmeq->iopoll);
>> +		return IRQ_HANDLED;
>> +	}
> 
> The disable_irq_nosync still does non-posted transactions for msi and
> msi-x. NVMe has a shortcut for msi, but there's currently nothing for
> msi-x. For that case, I have a prep patch I consulted with Thomas, and
> the result is below. I'm delayed sending it out while catching up from
> travel, but I think this will be even more efficient.
> 

Thanks for sharing. I'll give it a try. IIUC, after your patch goes in, 
I should be using disable_irq_nowait() in above and see better 
performance. Please correct me if I am wrong. I'll monitor your patch in 
LKML for future versions.

Regards,
Naman