On Mon, Nov 14, 2022 at 12:45:16PM -0800, Song Liu wrote: > On Sun, Nov 13, 2022 at 2:43 AM Mike Rapoport <rppt@xxxxxxxxxx> wrote: > > Actually, it would be interesting to quantify memory savings/waste as the > > result of using execmem_alloc() > > From a random system in our fleet, execmem_alloc() saves: > > 139 iTLB entries (1x 2MB entry vs, 140x 4kB entries), which is more than > 100% of L1 iTLB and about 10% of L2 TLB. That should be easily reflected then using perf using a real benchmark, however we don't have such data yet. I have hinted I suspect the reason might be that the actual loading of eBPF JIT / whatever may need to be done in parallel a la lib/test_kmod.c, or having the existing ebpf selftests run JIT tests in parallel, something along those lines, but I see no effort to showcase that. That's a missed opportunity. Luis
