On 7/15/26 5:45 PM, Muchun Song wrote: > >> On Jul 15, 2026, at 14:30, Li Zhe <[email protected]> wrote: >> >> On 7/14/26 7:22 PM, Muchun Song wrote: >>>> On Jul 14, 2026, at 17:45, Muchun Song <[email protected]> wrote: >>>> >>>> >>>> >>>> On 2026/7/9 19:25, Li Zhe wrote: >>>>> The template fast path currently uses memcpy() for the actual struct >>>>> page copy. Switch zone_device_page_init_from_template() to memcpy_nt() >>>>> and add memcpy_nt_drain() before memmap_init_compound(), before >>>>> prep_compound_head() updates overlapping tail metadata, and before >>>>> returning from memmap_init_zone_device(). >>>>> >>>>> ZONE_DEVICE memmap initialization is largely write-once: each struct >>>>> page is populated once, and most destination cachelines are not expected >>>>> to be reused immediately afterwards. On x86, a regular cached memcpy() >>>>> can therefore incur write-allocate traffic by pulling destination >>>>> cachelines into the cache before writeback, and can populate the cache >>>>> with data that has little near-term reuse. Using memcpy_nt() lets this >>>>> path request non-temporal stores for that copy pattern, which can reduce >>>>> cache pollution and avoid part of the associated write-allocate >>>>> overhead, while architectures without a specialized backend still fall >>>>> back to memcpy(). >>>>> >>>>> When memcpy_nt() maps to non-temporal stores, order those stores before >>>>> memmap_init_compound(), before prep_compound_head() updates overlapping >>>>> compound metadata, and before returning from memmap_init_zone_device(). >>>>> >>>>> Keep sanitized builds on the slow path so KASAN/KMSAN retain their >>>>> instrumented stores. >>>>> >>>>> Tested in a VM with a 100 GB fsdax namespace device configured with >>>>> map=dev and a 100 GB devdax namespace (align=2097152) on Intel Ice Lake >>>>> server. >>>>> >>>>> Test procedure: >>>>> Rebind the nd_pmem and dax_pmem driver 30 times and collect the memmap >>>>> initialization time from the pr_debug() output of >>>>> memmap_init_zone_device(). >>>>> >>>>> Base(v7.2-rc1): >>>>> First binding for nd_pmem driver: 1456 ms >>>>> Average of subsequent rebinds: 244.28 ms >>>>> >>>>> First binding for dax_pmem driver: 1462 ms >>>>> Average of subsequent rebinds: 273.31 ms >>>>> >>>>> With this series: >>>>> First binding for nd_pmem driver: 1272 ms >>>>> Average of subsequent rebinds: 96.79 ms >>>>> >>>>> First binding for dax_pmem driver: 1354 ms >>>>> Average of subsequent rebinds: 119.04 ms >>>>> >>>>> This reduces the average rebind time by about 60.4% for nd_pmem and >>>>> 56.4% for dax_pmem. >>>>> >>>>> Signed-off-by: Li Zhe <[email protected]> >>>>> --- >>>>> mm/mm_init.c | 39 +++++++++++++++++++++++++++++++++++++-- >>>>> 1 file changed, 37 insertions(+), 2 deletions(-) >>>>> >>>>> diff --git a/mm/mm_init.c b/mm/mm_init.c >>>>> index fb855bb0437a..addb4969587e 100644 >>>>> --- a/mm/mm_init.c >>>>> +++ b/mm/mm_init.c >>>>> @@ -1067,11 +1067,21 @@ static void __ref >>>>> zone_device_page_init_slow(struct page *page, >>>>> >>>>> static inline bool zone_device_page_init_optimization_enabled(void) >>>>> { >>>>> + /* >>>>> + * Keep sanitized builds on the slow path so their stores stay >>>>> + * instrumented. >>>>> + */ >>>>> + if (IS_ENABLED(CONFIG_KASAN) || IS_ENABLED(CONFIG_KMSAN)) >>>>> + return false; >>>> It is not a major concern if struct page initialization lacks KASAN or >>>> KMSAN instrumentation. To keep things simple, let's just entirely remove >>>> zone_device_page_init_optimization_enabled() to simplify the code. >>> BTW, If you really want to properly support KASAN, you should integrate >>> the check directly into memcpy_flushcache() via kasan_check_write(), rather >>> than treating it as a special case here. >> I agree that proper KASAN-aware support, if ever needed for >> memcpy_flushcache(), would belong in the low-level helper itself rather >> than in this caller. >> >> That said, for this series I would still prefer to keep the local check >> in zone_device_page_init_optimization_enabled(). The goal here is not to >> add full KASAN support to memcpy_flushcache(), but simply to keep >> sanitized builds on the existing instrumented slow path. >> >> When KASAN/KMSAN is enabled, performance is usually not the priority for >> that configuration, while preserving the instrumented store path is. >> Keeping the check local here avoids broadening the scope of this series >> into a generic memcpy_flushcache() change. > However, there is an even simpler approach: just don't do any special handling > for KASAN. In fact, many callers of memcpy_flushcache don't handle it at all. > > So I'm curious, why are you keen on handling KASAN as a special case here? > > Thanks. I agree that this is a simpler direction.
If we do not need to keep the sanitizer case on the non-template path, then there is little value in keeping a separate zone_device_page_init_optimization_enabled() helper either. The template path can be the only initialization path, and the old non-template helper code can be removed. If this looks good to everyone, I plan to do that as a separate cleanup patch in the next version, likely as patch 9. The reason for keeping it separate is that patches 4 and 5 introduce the template initialization logic, while patch 8 only switches the copy primitive used by that path. Dropping the remaining non-template fallback is a follow-up cleanup made possible by this discussion, so a separate patch should make the change easier to review: first add and use the template path, then remove the now-unneeded fallback code. Thanks, Zhe >> Thanks, >> Zhe >> >>>> Thanks. >>>> >>>>> + >>>>> /* >>>>> * The template fast path copies a preinitialized struct page image. >>>>> * Skip it when the page_ref_set tracepoint is enabled. >>>>> */ >>>>> - return !page_ref_tracepoint_active(page_ref_set); >>>>> + if (page_ref_tracepoint_active(page_ref_set)) >>>>> + return false; >>>>> + >>>>> + return true; >>>>> } >>>>> >>>>> static inline void zone_device_tail_page_init(struct page *page, >>>>> @@ -1110,7 +1120,7 @@ static void >>>>> zone_device_page_init_from_template(struct page *page, >>>>> * to the destination page. >>>>> */ >>>>> zone_device_page_update_template(template, pfn); >>>>> - memcpy(page, template, sizeof(*page)); >>>>> + memcpy_nt(page, template, sizeof(*page)); >>>>> } >>>>> >>>>> /* >>>>> @@ -1179,6 +1189,15 @@ static void __ref memmap_init_compound(struct page >>>>> *head, >>>>> &template); >>>>> } >>>>> } >>>>> + >>>>> + /* >>>>> + * When the template path is enabled, order the preceding tail-page >>>>> copies >>>>> + * before prep_compound_head() updates the overlapping compound metadata >>>>> + * in the first tail-page descriptors. If memcpy_nt() fell back to >>>>> + * regular cached stores, memcpy_nt_drain() may be a no-op. >>>>> + */ >>>>> + if (use_template) >>>>> + memcpy_nt_drain(); >>>>> prep_compound_head(head, order); >>>>> } >>>>> >>>>> @@ -1238,10 +1257,26 @@ void __ref memmap_init_zone_device(struct zone >>>>> *zone, >>>>> if (pfns_per_compound == 1) >>>>> continue; >>>>> >>>>> + /* >>>>> + * When the template path is enabled, order the preceding head-page copy >>>>> + * before memmap_init_compound(), which immediately updates compound-head >>>>> + * metadata. If memcpy_nt() fell back to regular cached stores, >>>>> + * memcpy_nt_drain() may be a no-op. >>>>> + */ >>>>> + if (use_template) >>>>> + memcpy_nt_drain(); >>>>> + >>>>> memmap_init_compound(page, pfn, zone_idx, nid, pgmap, >>>>> compound_nr_pages(pfn, altmap, pgmap), >>>>> use_template); >>>>> } >>>>> + /* >>>>> + * Ensure any prior template copies are ordered before returning. >>>>> + * On architectures where memcpy_nt() used regular cached stores, >>>>> + * memcpy_nt_drain() may be a no-op. >>>>> + */ >>>>> + if (use_template) >>>>> + memcpy_nt_drain(); >>>>> >>>>> pageblock_migratetype_init_range(start_pfn, nr_pages, MIGRATE_MOVABLE); >>>>> >>>>> -- >>>>> 2.20.1
