> On Nov 16, 2025, at 11:32 PM, Jason Wang <[email protected]> wrote: > > On Fri, Nov 14, 2025 at 10:53 PM Jon Kohler <[email protected]> wrote: >> >> >> >>> On Nov 12, 2025, at 8:09 PM, Jason Wang <[email protected]> wrote: >>> >>> !-------------------------------------------------------------------| >>> CAUTION: External Email >>> >>> |-------------------------------------------------------------------! >>> >>> On Thu, Nov 13, 2025 at 8:14 AM Jon Kohler <[email protected]> wrote: >>>> >>>> vhost_get_user and vhost_put_user leverage __get_user and __put_user, >>>> respectively, which were both added in 2016 by commit 6b1e6cc7855b >>>> ("vhost: new device IOTLB API"). >>> >>> It has been used even before this commit. >> >> Ah, thanks for the pointer. I’d have to go dig to find its genesis, but >> its more to say, this existed prior to the LFENCE commit. >> >>> >>>> In a heavy UDP transmit workload on a >>>> vhost-net backed tap device, these functions showed up as ~11.6% of >>>> samples in a flamegraph of the underlying vhost worker thread. >>>> >>>> Quoting Linus from [1]: >>>> Anyway, every single __get_user() call I looked at looked like >>>> historical garbage. [...] End result: I get the feeling that we >>>> should just do a global search-and-replace of the __get_user/ >>>> __put_user users, replace them with plain get_user/put_user instead, >>>> and then fix up any fallout (eg the coco code). >>>> >>>> Switch to plain get_user/put_user in vhost, which results in a slight >>>> throughput speedup. get_user now about ~8.4% of samples in flamegraph. >>>> >>>> Basic iperf3 test on a Intel 5416S CPU with Ubuntu 25.10 guest: >>>> TX: taskset -c 2 iperf3 -c <rx_ip> -t 60 -p 5200 -b 0 -u -i 5 >>>> RX: taskset -c 2 iperf3 -s -p 5200 -D >>>> Before: 6.08 Gbits/sec >>>> After: 6.32 Gbits/sec >>> >>> I wonder if we need to test on archs like ARM. >> >> Are you thinking from a performance perspective? Or a correctness one? > > Performance, I think the patch is correct. > > Thanks >
Ok gotcha. If anyone has an ARM system stuffed in their front pocket and can give this a poke, I’d appreciate it, as I don’t have ready access to one personally. That said, I think this might end up in “well, it is what it is” territory as Linus was alluding to, i.e. if performance dips on ARM for vhost, then thats a compelling point to optimize whatever ends up being the culprit for get/put user? Said another way, would ARM perf testing (or any other arch) be a blocker to taking this change? Thanks - Jon
