[Crash-utility] Re: [PATCH v3] "kmem -i" extremely slow on dumps from large memory systems

The "kmem -i" command is extremely slow (appears to hang) on dumps from large memory systems. For example, on 120GB crash dump from a high-end server with 32TB of RAM (ie. 8Giga Pages), the "kmem -i" command is taking over 50 minutes to execute on a DL380 Gen10. To report basic general memory usage figures, we should only be reading global counters, without having to walk the full flat/sparse mem map page table. Hence, dump_kmeminfo() should first be reading globals, and then only call dump_mem_map() if important information (ie. slabs or total ram) is missing. Signed-off-by: Georges Aureau <georges.aureau(a)hpe.com&gt; -- help.c | 4 +++- memory.c | 72 +++++++++++++++++++++++++++++++++++++++++++++++----------------- 2 files changed, 56 insertions(+), 20 deletions(-) diff --git a/help.c b/help.c index 4c30047..8bd74ac 100644 --- a/help.c +++ b/help.c @@ -6825,7 +6825,9 @@ char *help_kmem[] = { " -F same as -f, but also dumps all pages linked to that header.", " -c walks through the page_hash_table and verifies page_cache_size.", " -C same as -c, but also dumps all pages in the page_hash_table.", -" -i displays general memory usage information", +" -i displays general memory usage information.", +" Note: SHARED is no longer printed (too slow on large memory systems)", +" unless specifying \"-i=shared\".", " -v displays the mapped virtual memory regions allocated by vmalloc().", " -V displays the kernel vm_stat table if it exists, or in more recent", " kernels, the vm_zone_stat, vm_node_stat and vm_numa_stat tables,", diff --git a/memory.c b/memory.c index 8c01ed0..ea00ef1 100644 --- a/memory.c +++ b/memory.c @@ -237,7 +237,7 @@ static void dump_vmlist(struct meminfo *); static void dump_vmap_area(struct meminfo *); static int get_vmap_area_list_from_nodes(ulong **); static int dump_page_lists(struct meminfo *); -static void dump_kmeminfo(void); +static void dump_kmeminfo(struct meminfo *); static int page_to_phys(ulong, physaddr_t *); static void display_memory(ulonglong, long, ulong, int, void *); static char *show_opt_string(struct searchinfo *); @@ -5107,7 +5107,7 @@ cmd_kmem(void) BZERO(&value[0], sizeof(ulonglong)*MAXARGS); pc->curcmd_flags &= ~HEADER_PRINTED; - while ((c = getopt(argcnt, args, "gI:sS::rFfm:pvczCinl:L:PVoh")) != EOF) { + while ((c = getopt(argcnt, args, "gI:sS::rFfm:pvczCi::nl:L:PVoh")) != EOF) { switch(c) { case 'V': @@ -5124,6 +5124,9 @@ cmd_kmem(void) case 'i': iflag = 1; + if (optarg && strcmp(optarg, "=shared") == 0) { + meminfo.flags = GET_SHARED_PAGES; + } break; case 'h': @@ -5410,7 +5413,7 @@ cmd_kmem(void) } if (iflag == 1) - dump_kmeminfo(); + dump_kmeminfo(&meminfo); if (pflag == 1) dump_mem_map(&meminfo); @@ -8515,7 +8518,7 @@ bailout: char *kmeminfo_hdr = " PAGES TOTAL PERCENTAGE\n"; static void -dump_kmeminfo(void) +dump_kmeminfo(struct meminfo *mi) { int i, len; ulong totalram_pages; @@ -8546,18 +8549,23 @@ dump_kmeminfo(void) ulong get_buffers; ulong get_slabs; char buf[BUFSIZE]; + ulong flags; - - BZERO(&meminfo, sizeof(struct meminfo)); - meminfo.flags = GET_ALL; - dump_mem_map(&meminfo); - get_totalram = meminfo.get_totalram; - shared_pages = meminfo.get_shared; - get_buffers = meminfo.get_buffers; - get_slabs = meminfo.get_slabs; + /* + * By default, we will no longer call dump_mem_map() as this is too + * slow for large memory systems. If we have to call it (eg. missing + * important information such as slabs or total ram), we will also + * collect shared pages. Otherwise, we won't print shared pages unless + * the caller explicitly requested shared pages ("kmem -i=shared"). + */ + flags = mi->flags; + shared_pages = 0; + get_totalram = 0; + get_buffers = 0; + get_slabs = 0; /* - * If vm_stat array exists, override page search info. + * If vm_stat array does not exists, then set mem map flag. */ if (vm_stat_init()) { if (dump_vm_stat("NR_SLAB", &nr_slab, 0)) @@ -8572,9 +8580,10 @@ dump_kmeminfo(void) if (dump_vm_stat("NR_SLAB_UNRECLAIMABLE_B", &nr_slab, 0)) get_slabs += nr_slab; } + } else { + flags |= GET_SLAB_PAGES; } - fprintf(fp, "%s", kmeminfo_hdr); /* * Get total RAM based upon how the various versions of si_meminfo() * have done it, latest to earliest: @@ -8586,9 +8595,32 @@ dump_kmeminfo(void) symbol_exists("_totalram_pages")) { totalram_pages = vt->totalram_pages ? vt->totalram_pages : get_totalram; - } else - totalram_pages = get_totalram; + } else { + flags |= GET_TOTALRAM_PAGES; + totalram_pages = 0; + } + /* + * If the caller wants shared pages or if we are missing important data + * (ie. slab or totalram) then go through the slow dump_mem_map() path. + */ + if (flags) { + BZERO(&meminfo, sizeof(struct meminfo)); + meminfo.flags = GET_ALL; + dump_mem_map(&meminfo); + /* Update the missing information */ + if (flags & GET_SLAB_PAGES) { + get_slabs = meminfo.get_slabs; + } + if (flags & GET_TOTALRAM_PAGES) { + get_totalram = meminfo.get_totalram; + totalram_pages = get_totalram; + } + shared_pages = meminfo.get_shared; + get_buffers = meminfo.get_buffers; + } + + fprintf(fp, "%s", kmeminfo_hdr); fprintf(fp, "%13s %7ld %11s ----\n", "TOTAL MEM", totalram_pages, pages_to_size(totalram_pages, buf)); @@ -8613,9 +8645,11 @@ dump_kmeminfo(void) * differently than the kernel -- it just tallies the non-reserved * pages that have a count of greater than 1. */ - pct = (shared_pages * 100)/totalram_pages; - fprintf(fp, "%13s %7ld %11s %3ld%% of TOTAL MEM\n", - "SHARED", shared_pages, pages_to_size(shared_pages, buf), pct); + if (flags & GET_SHARED_PAGES) { + pct = (shared_pages * 100)/totalram_pages; + fprintf(fp, "%13s %7ld %11s %3ld%% of TOTAL MEM\n", + "SHARED", shared_pages, pages_to_size(shared_pages, buf), pct); + } subtract_buffer_pages = 0; if (symbol_exists("buffermem_pages")) { -- Crash-utility mailing list -- devel(a)lists.crash-utility.osci.io To unsubscribe send an email to devel-leave(a)lists.crash-utility.osci.io https://${domain_name}/admin/lists/devel.lists.crash-utility.osci.io/ Contribution Guidelines: https://github.com/crash-utility/crash/wiki