From 228957fea3a998735524abf6354634f1fb710e61 Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Thu, 8 Oct 2020 10:30:26 +0200 Subject: [PATCH] virtio-mem: Probe THP size to determine default block size Let's allow a minimum block size of 1 MiB in all configurations. Select the default block size based on - The page size of the memory backend. - The THP size if the memory backend size corresponds to the real host page size. - The global minimum of 1 MiB. and warn if something smaller is configured by the user. VIRTIO_MEM only supports Linux (depends on LINUX), so we can probe the THP size unconditionally. For now we only support virtio-mem on x86-64 - there isn't a user-visible change (x86-64 only supports 2 MiB THP on the PMD level) - the default was, and will be 2 MiB. If we ever have THP on the PUD level (e.g., 1 GiB THP on x86-64), we expect it to be more transparent - e.g., to only optimize fully populated ranges unless explicitly told /configured otherwise (in contrast to PMD THP). Reviewed-by: Pankaj Gupta Cc: "Michael S. Tsirkin" Cc: Wei Yang Cc: Dr. David Alan Gilbert Cc: Igor Mammedov Cc: Pankaj Gupta Signed-off-by: David Hildenbrand Message-Id: <20201008083029.9504-4-david@redhat.com> Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/virtio/virtio-mem.c | 105 +++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 101 insertions(+), 4 deletions(-) diff --git a/hw/virtio/virtio-mem.c b/hw/virtio/virtio-mem.c index 461ac68ee8..655824ff81 100644 --- a/hw/virtio/virtio-mem.c +++ b/hw/virtio/virtio-mem.c @@ -33,10 +33,83 @@ #include "trace.h" /* - * Use QEMU_VMALLOC_ALIGN, so no THP will have to be split when unplugging - * memory (e.g., 2MB on x86_64). + * Let's not allow blocks smaller than 1 MiB, for example, to keep the tracking + * bitmap small. */ -#define VIRTIO_MEM_MIN_BLOCK_SIZE ((uint32_t)QEMU_VMALLOC_ALIGN) +#define VIRTIO_MEM_MIN_BLOCK_SIZE ((uint32_t)(1 * MiB)) + +#if defined(__x86_64__) || defined(__arm__) || defined(__aarch64__) || \ + defined(__powerpc64__) +#define VIRTIO_MEM_DEFAULT_THP_SIZE ((uint32_t)(2 * MiB)) +#else + /* fallback to 1 MiB (e.g., the THP size on s390x) */ +#define VIRTIO_MEM_DEFAULT_THP_SIZE VIRTIO_MEM_MIN_BLOCK_SIZE +#endif + +/* + * We want to have a reasonable default block size such that + * 1. We avoid splitting THPs when unplugging memory, which degrades + * performance. + * 2. We avoid placing THPs for plugged blocks that also cover unplugged + * blocks. + * + * The actual THP size might differ between Linux kernels, so we try to probe + * it. In the future (if we ever run into issues regarding 2.), we might want + * to disable THP in case we fail to properly probe the THP size, or if the + * block size is configured smaller than the THP size. + */ +static uint32_t thp_size; + +#define HPAGE_PMD_SIZE_PATH "/sys/kernel/mm/transparent_hugepage/hpage_pmd_size" +static uint32_t virtio_mem_thp_size(void) +{ + gchar *content = NULL; + const char *endptr; + uint64_t tmp; + + if (thp_size) { + return thp_size; + } + + /* + * Try to probe the actual THP size, fallback to (sane but eventually + * incorrect) default sizes. + */ + if (g_file_get_contents(HPAGE_PMD_SIZE_PATH, &content, NULL, NULL) && + !qemu_strtou64(content, &endptr, 0, &tmp) && + (!endptr || *endptr == '\n')) { + /* + * Sanity-check the value, if it's too big (e.g., aarch64 with 64k base + * pages) or weird, fallback to something smaller. + */ + if (!tmp || !is_power_of_2(tmp) || tmp > 16 * MiB) { + warn_report("Read unsupported THP size: %" PRIx64, tmp); + } else { + thp_size = tmp; + } + } + + if (!thp_size) { + thp_size = VIRTIO_MEM_DEFAULT_THP_SIZE; + warn_report("Could not detect THP size, falling back to %" PRIx64 + " MiB.", thp_size / MiB); + } + + g_free(content); + return thp_size; +} + +static uint64_t virtio_mem_default_block_size(RAMBlock *rb) +{ + const uint64_t page_size = qemu_ram_pagesize(rb); + + /* We can have hugetlbfs with a page size smaller than the THP size. */ + if (page_size == qemu_real_host_page_size) { + return MAX(page_size, virtio_mem_thp_size()); + } + return MAX(page_size, VIRTIO_MEM_MIN_BLOCK_SIZE); +} + /* * Size the usable region bigger than the requested size if possible. Esp. * Linux guests will only add (aligned) memory blocks in case they fully @@ -443,10 +516,23 @@ static void virtio_mem_device_realize(DeviceState *dev, Error **errp) rb = vmem->memdev->mr.ram_block; page_size = qemu_ram_pagesize(rb); + /* + * If the block size wasn't configured by the user, use a sane default. This + * allows using hugetlbfs backends of any page size without manual + * intervention. + */ + if (!vmem->block_size) { + vmem->block_size = virtio_mem_default_block_size(rb); + } + if (vmem->block_size < page_size) { error_setg(errp, "'%s' property has to be at least the page size (0x%" PRIx64 ")", VIRTIO_MEM_BLOCK_SIZE_PROP, page_size); return; + } else if (vmem->block_size < virtio_mem_default_block_size(rb)) { + warn_report("'%s' property is smaller than the default block size (%" + PRIx64 " MiB)", VIRTIO_MEM_BLOCK_SIZE_PROP, + virtio_mem_default_block_size(rb) / MiB); } else if (!QEMU_IS_ALIGNED(vmem->requested_size, vmem->block_size)) { error_setg(errp, "'%s' property has to be multiples of '%s' (0x%" PRIx64 ")", VIRTIO_MEM_REQUESTED_SIZE_PROP, @@ -742,6 +828,18 @@ static void virtio_mem_get_block_size(Object *obj, Visitor *v, const char *name, const VirtIOMEM *vmem = VIRTIO_MEM(obj); uint64_t value = vmem->block_size; + /* + * If not configured by the user (and we're not realized yet), use the + * default block size we would use with the current memory backend. + */ + if (!value) { + if (vmem->memdev && memory_region_is_ram(&vmem->memdev->mr)) { + value = virtio_mem_default_block_size(vmem->memdev->mr.ram_block); + } else { + value = virtio_mem_thp_size(); + } + } + visit_type_size(v, name, &value, errp); } @@ -821,7 +919,6 @@ static void virtio_mem_instance_init(Object *obj) { VirtIOMEM *vmem = VIRTIO_MEM(obj); - vmem->block_size = VIRTIO_MEM_MIN_BLOCK_SIZE; notifier_list_init(&vmem->size_change_notifiers); vmem->precopy_notifier.notify = virtio_mem_precopy_notify; -- 2.11.0