On Thu, 16 Mar 2023 at 16:10, Zhenhua Huang <quic_zhenhuah@xxxxxxxxxxx> wrote: > > Kfence only needs its pool to be mapped as page granularity, if it is > inited early. Previous judgement was a bit over protected. From [1], Mark > suggested to "just map the KFENCE region a page granularity". So I > decouple it from judgement and do page granularity mapping for kfence > pool only. Need to be noticed that late init of kfence pool still requires > page granularity mapping. > > Page granularity mapping in theory cost more(2M per 1GB) memory on arm64 > platform. Like what I've tested on QEMU(emulated 1GB RAM) with > gki_defconfig, also turning off rodata protection: > Before: > [root@liebao ]# cat /proc/meminfo > MemTotal: 999484 kB > After: > [root@liebao ]# cat /proc/meminfo > MemTotal: 1001480 kB > > To implement this, also relocate the kfence pool allocation before the > linear mapping setting up, arm64_kfence_alloc_pool is to allocate phys > addr, __kfence_pool is to be set after linear mapping set up. > > LINK: [1] https://lore.kernel.org/linux-arm-kernel/Y+IsdrvDNILA59UN@FVFF77S0Q05N/ > Suggested-by: Mark Rutland <mark.rutland@xxxxxxx> > Signed-off-by: Zhenhua Huang <quic_zhenhuah@xxxxxxxxxxx> > Reviewed-by: Kefeng Wang <wangkefeng.wang@xxxxxxxxxx> Reviewed-by: Marco Elver <elver@xxxxxxxxxx> One question: what happens if the page-granular direct map is requested either way, is there any downside with this patch? Does it mean map_mem() just does a little extra work it shouldn't have? (Not saying that's a problem, just trying to ask you to double-check it's ok.) However, please also wait for an arm64 maintainer to have a look. I'm assuming that because it touches mostly arm64 code, this patch ought to go through the arm64 tree? Thanks, -- Marco
