From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Gleixner Date: Thu, 22 Oct 2020 10:28:02 +0200 Subject: [Intel-wired-lan] [PATCH v4 4/4] PCI: Limit pci_alloc_irq_vectors() to housekeeping CPUs In-Reply-To: <20201021170224.55aea948@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com> References: <20200928183529.471328-1-nitesh@redhat.com> <20200928183529.471328-5-nitesh@redhat.com> <87v9f57zjf.fsf@nanos.tec.linutronix.de> <3bca9eb1-a318-1fc6-9eee-aacc0293a193@redhat.com> <87lfg093fo.fsf@nanos.tec.linutronix.de> <877drj72cz.fsf@nanos.tec.linutronix.de> <20201021170224.55aea948@kicinski-fedora-pc1c0hjn.dhcp.thefacebook.com> Message-ID: <874kmm7jhp.fsf@nanos.tec.linutronix.de> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: intel-wired-lan@osuosl.org List-ID: On Wed, Oct 21 2020 at 17:02, Jakub Kicinski wrote: > On Wed, 21 Oct 2020 22:25:48 +0200 Thomas Gleixner wrote: >> The right answer to this is to utilize managed interrupts and have >> according logic in your network driver to handle CPU hotplug. When a CPU >> goes down, then the queue which is associated to that CPU is quiesced >> and the interrupt core shuts down the relevant interrupt instead of >> moving it to an online CPU (which causes the whole vector exhaustion >> problem on x86). When the CPU comes online again, then the interrupt is >> reenabled in the core and the driver reactivates the queue. > > I think Mellanox folks made some forays into managed irqs, but I don't > remember/can't find the details now. > > For networking the locality / queue per core does not always work, > since the incoming traffic is usually spread based on a hash. Many That makes it problematic and is fundamentally different from block I/O. > applications perform better when network processing is done on a small > subset of CPUs, and application doesn't get interrupted every 100us. > So we do need extra user control here. Ok. > We have a bit of a uAPI problem since people had grown to depend on > IRQ == queue == NAPI to configure their systems. "The right way" out > would be a proper API which allows associating queues with CPUs rather > than IRQs, then we can use managed IRQs and solve many other problems. > > Such new API has been in the works / discussions for a while now. If there is anything which needs to be done/extended on the irq side please let me know. Thanks tglx