From patchwork Mon Sep 5 08:15:49 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Sergey Senozhatsky X-Patchwork-Id: 12965753 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id 4B011ECAAD5 for ; Mon, 5 Sep 2022 08:16:25 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id DC09D801C0; Mon, 5 Sep 2022 04:16:24 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id D49C6801B3; Mon, 5 Sep 2022 04:16:24 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id B9BF8801C0; Mon, 5 Sep 2022 04:16:24 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0010.hostedemail.com [216.40.44.10]) by kanga.kvack.org (Postfix) with ESMTP id A5847801B3 for ; Mon, 5 Sep 2022 04:16:24 -0400 (EDT) Received: from smtpin27.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay01.hostedemail.com (Postfix) with ESMTP id 703D61C5D9D for ; Mon, 5 Sep 2022 08:16:24 +0000 (UTC) X-FDA: 79877324688.27.D85D6E6 Received: from mail-pg1-f171.google.com (mail-pg1-f171.google.com [209.85.215.171]) by imf14.hostedemail.com (Postfix) with ESMTP id 334E7100052 for ; Mon, 5 Sep 2022 08:16:24 +0000 (UTC) Received: by mail-pg1-f171.google.com with SMTP id b196so7495428pga.7 for ; Mon, 05 Sep 2022 01:16:23 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date; bh=apGSrKTgwETbH2CMEGMydpnHgs87JcMdkEf7eiEN8iQ=; b=UHiLRbvmr56hCocfhGGKjJpCVX0QmwXKnDK02ZfwGPiXbrffOhTCeCLMKNpR9/tYGN X3yaRX3dSDUu5HLs3OpqRGVnskUagYD2WNBKNj5vf5SlhMZttWoROgu0ZSCMWH0M/cgf SvFJoxTLExmSdSrU8YjuphWxZVJHzqbq7m4H8= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date; bh=apGSrKTgwETbH2CMEGMydpnHgs87JcMdkEf7eiEN8iQ=; b=xyjmfd1vm+JLU7Ug6YBoH9lbVztvKG85tZcnppdv2SWVTij5haMobI24vgNTRJIHky e4quSll313sXfpNszKDNlX1ZlYOiNLsyA/ZUdcwTfJ1OBc7dWeZ1AMLt3PjBBwX5o/47 /i8kNQjECk7CxAKkBTLkaKl5YyCld+KlOFYNv7URUKWtoM/K0pkL8WMCfUr3xQK6dp4L A/KJeiqwk20kAc4R51ngFdF1yWRG1rVigtzMo/6UucQqzelowxOGZJxHaUobW54KQQyh 4Vfq6sClN+CgGSxAhHbDRGEn/ykooK8dQ47kjyQbpuy/t4jNb2/MjzN+MgvO5pASWl2O /80g== X-Gm-Message-State: ACgBeo2cMX98TsMtnrTcWKHsnVe2uy8M+2iAyUFJqf3JmNM4DJphRLdj B6DNGMa053SocOLdmDPSfHHfLw== X-Google-Smtp-Source: AA6agR5FiXtZ9BmQX6KCmb1jFq9Sn1Jf93cqLIli9ZyWSjqB65afzwKK1weUMhLUAwMqp9lmZDgNPQ== X-Received: by 2002:a05:6a00:1907:b0:534:f2ef:e8fe with SMTP id y7-20020a056a00190700b00534f2efe8femr48844002pfi.65.1662365783216; Mon, 05 Sep 2022 01:16:23 -0700 (PDT) Received: from tigerii.tok.corp.google.com ([2401:fa00:8f:203:71e2:8444:42d9:4bb0]) by smtp.gmail.com with ESMTPSA id w12-20020aa79a0c000000b005363bc65bafsm7075864pfj.57.2022.09.05.01.16.21 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Mon, 05 Sep 2022 01:16:22 -0700 (PDT) From: Sergey Senozhatsky To: Minchan Kim , Andrew Morton Cc: Nitin Gupta , linux-kernel@vger.kernel.org, linux-mm@kvack.org, Sergey Senozhatsky Subject: [PATCH RFC 4/7] zram: Introduce recompress sysfs knob Date: Mon, 5 Sep 2022 17:15:49 +0900 Message-Id: <20220905081552.2740917-8-senozhatsky@chromium.org> X-Mailer: git-send-email 2.37.2.789.g6183377224-goog In-Reply-To: <20220905081552.2740917-1-senozhatsky@chromium.org> References: <20220905081552.2740917-1-senozhatsky@chromium.org> MIME-Version: 1.0 ARC-Authentication-Results: i=1; imf14.hostedemail.com; dkim=pass header.d=chromium.org header.s=google header.b=UHiLRbvm; spf=pass (imf14.hostedemail.com: domain of senozhatsky@chromium.org designates 209.85.215.171 as permitted sender) smtp.mailfrom=senozhatsky@chromium.org; dmarc=pass (policy=none) header.from=chromium.org ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1662365784; a=rsa-sha256; cv=none; b=G7dGbAtlKeU6FK+PGeuO7xQcLZEy5M4FB1UOnUv8s+zYOkDgtwW7fTZeBuGfrK9gVUYqt+ uRGVt77Gm1sWTWj3fBThdFEUL7U33nFoYiMdJrR/XAtNnBLM5VftFyWqA6uMjV9a+32526 tA6zITpLCL8421kWRIr4lwvf+o1h6hk= ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1662365784; h=from:from:sender:reply-to:subject:subject:date:date: message-id:message-id:to:to:cc:cc:mime-version:mime-version: content-type:content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references:dkim-signature; bh=apGSrKTgwETbH2CMEGMydpnHgs87JcMdkEf7eiEN8iQ=; b=3n7Fu7DWmgZuk4M1jbkllU32HL91qXZtEI2u3ppQmaBCCkINpDKKiS4cKyZM6lF/hTb/CO 863HJcr6n03rFtISG+tzLp83XkxaIhGsvphWi2q7vmHyT5cVFv5M1+uaUQXTuN6fyYzbmX MY94T3ZUlRCq9jCsHtLmTkN6DUfz3vY= X-Stat-Signature: xa6c6c6i7emk1kk84jofuty37jj68e8w X-Rspamd-Queue-Id: 334E7100052 X-Rspam-User: Authentication-Results: imf14.hostedemail.com; dkim=pass header.d=chromium.org header.s=google header.b=UHiLRbvm; spf=pass (imf14.hostedemail.com: domain of senozhatsky@chromium.org designates 209.85.215.171 as permitted sender) smtp.mailfrom=senozhatsky@chromium.org; dmarc=pass (policy=none) header.from=chromium.org X-Rspamd-Server: rspam07 X-HE-Tag: 1662365784-858254 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: Allow zram to recompress (using secondary compression streams) pages. We support three modes: 1) IDLE pages recompression is activated by `idle` mode echo idle > /sys/block/zram0/recompress 2) Since there may be many idle pages user-space may pass a size watermark value and we will recompress IDLE pages only of equal or greater size: echo 888 > /sys/block/zram0/recompress 3) HUGE pages recompression is activated by `huge` mode echo huge > /sys/block/zram0/recompress 4) HUGE_IDLE pages recompression is activated by `huge_idle` mode echo huge_idle > /sys/block/zram0/recompress Signed-off-by: Sergey Senozhatsky --- drivers/block/zram/Kconfig | 11 ++ drivers/block/zram/zram_drv.c | 191 +++++++++++++++++++++++++++++++++- drivers/block/zram/zram_drv.h | 1 + 3 files changed, 200 insertions(+), 3 deletions(-) diff --git a/drivers/block/zram/Kconfig b/drivers/block/zram/Kconfig index d4100b0c083e..81ae4b96ec1a 100644 --- a/drivers/block/zram/Kconfig +++ b/drivers/block/zram/Kconfig @@ -78,3 +78,14 @@ config ZRAM_MEMORY_TRACKING /sys/kernel/debug/zram/zramX/block_state. See Documentation/admin-guide/blockdev/zram.rst for more information. + +config ZRAM_MULTI_COMP + bool "Enable multiple per-CPU compression streams" + depends on ZRAM + help + This will enable per-CPU multi-compression streams, so that ZRAM + can re-compress IDLE pages, using a potentially slower but more + effective compression algorithm. + + echo TIMEOUT > /sys/block/zramX/idle + echo SIZE > /sys/block/zramX/recompress diff --git a/drivers/block/zram/zram_drv.c b/drivers/block/zram/zram_drv.c index de2970865b7b..386e49a13806 100644 --- a/drivers/block/zram/zram_drv.c +++ b/drivers/block/zram/zram_drv.c @@ -1293,6 +1293,9 @@ static void zram_free_page(struct zram *zram, size_t index) atomic64_dec(&zram->stats.huge_pages); } + if (zram_test_flag(zram, index, ZRAM_RECOMP)) + zram_clear_flag(zram, index, ZRAM_RECOMP); + if (zram_test_flag(zram, index, ZRAM_WB)) { zram_clear_flag(zram, index, ZRAM_WB); free_block_bdev(zram, zram_get_element(zram, index)); @@ -1357,6 +1360,7 @@ static int zram_read_from_zspool(struct zram *zram, unsigned long handle; unsigned int size; void *src, *dst; + u32 idx; int ret; handle = zram_get_handle(zram, index); @@ -1373,8 +1377,13 @@ static int zram_read_from_zspool(struct zram *zram, size = zram_get_obj_size(zram, index); - if (size != PAGE_SIZE) - zstrm = zcomp_stream_get(zram->comps[ZRAM_PRIMARY_ZCOMP]); + if (size != PAGE_SIZE) { + idx = ZRAM_PRIMARY_ZCOMP; + if (zram_test_flag(zram, index, ZRAM_RECOMP)) + idx = ZRAM_SECONDARY_ZCOMP; + + zstrm = zcomp_stream_get(zram->comps[idx]); + } src = zs_map_object(zram->mem_pool, handle, ZS_MM_RO); if (size == PAGE_SIZE) { @@ -1386,7 +1395,7 @@ static int zram_read_from_zspool(struct zram *zram, dst = kmap_atomic(page); ret = zcomp_decompress(zstrm, src, size, dst); kunmap_atomic(dst); - zcomp_stream_put(zram->comps[ZRAM_PRIMARY_ZCOMP]); + zcomp_stream_put(zram->comps[idx]); } zs_unmap_object(zram->mem_pool, handle); return ret; @@ -1612,6 +1621,180 @@ static int zram_bvec_write(struct zram *zram, struct bio_vec *bvec, return ret; } +#ifdef CONFIG_ZRAM_MULTI_COMP +/* + * This function will decompress (unless it's ZRAM_HUGE) the page and then + * attempt to compress it using secondary compression algorithm (which is + * potentially more effective). + * + * Corresponding ZRAM slot should be locked. + */ +static int zram_recompress(struct zram *zram, + u32 index, + struct page *page, + int size_watermark) +{ + unsigned long handle_prev; + unsigned long handle_next; + unsigned int comp_len_next; + unsigned int comp_len_prev; + struct zcomp_strm *zstrm; + void *src, *dst; + int ret; + + handle_prev = zram_get_handle(zram, index); + if (!handle_prev) + return -EINVAL; + + comp_len_prev = zram_get_obj_size(zram, index); + /* + * Do not recompress objects that are already "small enough". + */ + if (comp_len_prev < size_watermark) + return 0; + + ret = zram_read_from_zspool(zram, page, index); + if (ret) + return ret; + + zstrm = zcomp_stream_get(zram->comps[ZRAM_SECONDARY_ZCOMP]); + src = kmap_atomic(page); + ret = zcomp_compress(zstrm, src, &comp_len_next); + kunmap_atomic(src); + + /* + * Either a compression error or we failed to compressed the object + * in a way that will save us memory. Mark object as "recompressed" + * if it's huge, so that we don't try to recompress it again. Ideally + * we want to set some bit for all such objects, but we for now do so + * only for huge ones (we are out of bits in flags on 32-bit systems). + */ + if (comp_len_next >= huge_class_size || + comp_len_next >= comp_len_prev || + ret) { + if (zram_test_flag(zram, index, ZRAM_HUGE)) + zram_set_flag(zram, index, ZRAM_RECOMP); + zram_clear_flag(zram, index, ZRAM_IDLE); + zcomp_stream_put(zram->comps[ZRAM_SECONDARY_ZCOMP]); + return ret; + } + + /* + * No direct reclaim (slow path) for handle allocation and no + * re-compression attempt (unlike in __zram_bvec_write()) since + * we already stored that object in zsmalloc. If we cannot alloc + * memory then me bail out. + */ + handle_next = zs_malloc(zram->mem_pool, comp_len_next, + __GFP_KSWAPD_RECLAIM | + __GFP_NOWARN | + __GFP_HIGHMEM | + __GFP_MOVABLE); + if (IS_ERR((void *)handle_next)) { + zcomp_stream_put(zram->comps[ZRAM_SECONDARY_ZCOMP]); + return -ENOMEM; + } + + dst = zs_map_object(zram->mem_pool, handle_next, ZS_MM_WO); + memcpy(dst, zstrm->buffer, comp_len_next); + zcomp_stream_put(zram->comps[ZRAM_SECONDARY_ZCOMP]); + + zs_unmap_object(zram->mem_pool, handle_next); + + zram_free_page(zram, index); + zram_set_handle(zram, index, handle_next); + zram_set_obj_size(zram, index, comp_len_next); + + zram_set_flag(zram, index, ZRAM_RECOMP); + atomic64_add(comp_len_next, &zram->stats.compr_data_size); + atomic64_inc(&zram->stats.pages_stored); + + return 0; +} + +#define RECOMPRESS_IDLE (1 << 0) +#define RECOMPRESS_HUGE (1 << 1) + +static ssize_t recompress_store(struct device *dev, + struct device_attribute *attr, + const char *buf, + size_t len) +{ + struct zram *zram = dev_to_zram(dev); + unsigned long nr_pages = zram->disksize >> PAGE_SHIFT; + unsigned long index; + struct page *page; + ssize_t ret = 0; + int mode, size_watermark = 0; + + if (sysfs_streq(buf, "idle")) { + mode = RECOMPRESS_IDLE; + } else if (sysfs_streq(buf, "huge")) { + mode = RECOMPRESS_HUGE; + } else if (sysfs_streq(buf, "huge_idle")) { + mode = RECOMPRESS_IDLE | RECOMPRESS_HUGE; + } else { + /* + * We will re-compress only idle objects equal or greater + * in size than watermark. + */ + ret = kstrtoint(buf, 10, &size_watermark); + if (ret) + return ret; + mode = RECOMPRESS_IDLE; + } + + if (size_watermark > PAGE_SIZE) + return ret; + + down_read(&zram->init_lock); + if (!init_done(zram)) { + ret = -EINVAL; + goto release_init_lock; + } + + page = alloc_page(GFP_KERNEL); + if (!page) { + ret = -ENOMEM; + goto release_init_lock; + } + + for (index = 0; index < nr_pages; index++) { + zram_slot_lock(zram, index); + + if (!zram_allocated(zram, index)) + goto next; + + if (mode & RECOMPRESS_IDLE && + !zram_test_flag(zram, index, ZRAM_IDLE)) + goto next; + + if (mode & RECOMPRESS_HUGE && + !zram_test_flag(zram, index, ZRAM_HUGE)) + goto next; + + if (zram_test_flag(zram, index, ZRAM_WB) || + zram_test_flag(zram, index, ZRAM_UNDER_WB) || + zram_test_flag(zram, index, ZRAM_SAME) || + zram_test_flag(zram, index, ZRAM_RECOMP)) + goto next; + + ret = zram_recompress(zram, index, page, size_watermark); +next: + zram_slot_unlock(zram, index); + if (ret) + break; + } + + ret = len; + __free_page(page); + +release_init_lock: + up_read(&zram->init_lock); + return ret; +} +#endif + /* * zram_bio_discard - handler on discard request * @index: physical block index in PAGE_SIZE units @@ -2001,6 +2184,7 @@ static DEVICE_ATTR_RW(writeback_limit_enable); #endif #ifdef CONFIG_ZRAM_MULTI_COMP static DEVICE_ATTR_RW(recomp_algorithm); +static DEVICE_ATTR_WO(recompress); #endif static struct attribute *zram_disk_attrs[] = { @@ -2027,6 +2211,7 @@ static struct attribute *zram_disk_attrs[] = { &dev_attr_debug_stat.attr, #ifdef CONFIG_ZRAM_MULTI_COMP &dev_attr_recomp_algorithm.attr, + &dev_attr_recompress.attr, #endif NULL, }; diff --git a/drivers/block/zram/zram_drv.h b/drivers/block/zram/zram_drv.h index af3d6f6bfcff..b4eecef2a11f 100644 --- a/drivers/block/zram/zram_drv.h +++ b/drivers/block/zram/zram_drv.h @@ -50,6 +50,7 @@ enum zram_pageflags { ZRAM_UNDER_WB, /* page is under writeback */ ZRAM_HUGE, /* Incompressible page */ ZRAM_IDLE, /* not accessed page since last idle marking */ + ZRAM_RECOMP, /* page was recompressed */ __NR_ZRAM_PAGEFLAGS, };