From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-9.8 required=3.0 tests=DKIM_INVALID,DKIM_SIGNED, INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS, USER_AGENT_GIT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 77B97C433E3 for ; Mon, 6 Jul 2020 17:21:22 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 2D31A206CD for ; Mon, 6 Jul 2020 17:21:22 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (1024-bit key) header.d=kernel.org header.i=@kernel.org header.b="yynfct9R" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 2D31A206CD Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=kernel.org Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id C84AF6B000A; Mon, 6 Jul 2020 13:21:21 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id C365C6B000C; Mon, 6 Jul 2020 13:21:21 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id B4B416B000D; Mon, 6 Jul 2020 13:21:21 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0053.hostedemail.com [216.40.44.53]) by kanga.kvack.org (Postfix) with ESMTP id A050F6B000A for ; Mon, 6 Jul 2020 13:21:21 -0400 (EDT) Received: from smtpin01.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay04.hostedemail.com (Postfix) with ESMTP id 5F1051EE6 for ; Mon, 6 Jul 2020 17:21:21 +0000 (UTC) X-FDA: 77008317162.01.beef45_2f0432526ead Received: from filter.hostedemail.com (10.5.16.251.rfc1918.com [10.5.16.251]) by smtpin01.hostedemail.com (Postfix) with ESMTP id C919640001CD57BC for ; Mon, 6 Jul 2020 17:21:20 +0000 (UTC) X-HE-Tag: beef45_2f0432526ead X-Filterd-Recvd-Size: 7849 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by imf50.hostedemail.com (Postfix) with ESMTP for ; Mon, 6 Jul 2020 17:21:20 +0000 (UTC) Received: from aquarius.haifa.ibm.com (nesher1.haifa.il.ibm.com [195.110.40.7]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-SHA256 (128/128 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id 188D420720; Mon, 6 Jul 2020 17:21:14 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1594056079; bh=HnBXNgi2qMQGGh+5sWPUWq04Tsq65N/aqIq2MmaYGEs=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=yynfct9RCgwmmU1XhrFRra9IagOCufD4Px41dHNiova2I/xRUkMyXFOK0+joi7k94 qIEWkpYeutLh46JhxonyGVN93bjvy18GdpD3nCIUr2rDqtq9Vp11u97as907YQv9Rb q/0QuRvezXKamERIQ9oqBX1fY0VUsllVziGet1Jc= From: Mike Rapoport To: linux-kernel@vger.kernel.org Cc: Alan Cox , Andrew Morton , Andy Lutomirski , Christopher Lameter , Dave Hansen , Idan Yaniv , James Bottomley , "Kirill A. Shutemov" , Matthew Wilcox , Peter Zijlstra , "Reshetova, Elena" , Thomas Gleixner , Tycho Andersen , linux-api@vger.kernel.org, linux-mm@kvack.org, Mike Rapoport Subject: [RFC PATCH v2 4/5] mm: secretmem: use PMD-size pages to amortize direct map fragmentation Date: Mon, 6 Jul 2020 20:20:50 +0300 Message-Id: <20200706172051.19465-5-rppt@kernel.org> X-Mailer: git-send-email 2.26.2 In-Reply-To: <20200706172051.19465-1-rppt@kernel.org> References: <20200706172051.19465-1-rppt@kernel.org> MIME-Version: 1.0 X-Rspamd-Queue-Id: C919640001CD57BC X-Spamd-Result: default: False [0.00 / 100.00] X-Rspamd-Server: rspam03 Content-Transfer-Encoding: quoted-printable 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: From: Mike Rapoport Removing a PAGE_SIZE page from the direct map every time such page is allocated for a secret memory mapping will cause severe fragmentation of the direct map. This fragmentation can be reduced by using PMD-size pages as a pool for small pages for secret memory mappings. Add a gen_pool per secretmem inode and lazily populate this pool with PMD-size pages. Signed-off-by: Mike Rapoport --- mm/secretmem.c | 107 ++++++++++++++++++++++++++++++++++++++++--------- 1 file changed, 88 insertions(+), 19 deletions(-) diff --git a/mm/secretmem.c b/mm/secretmem.c index df8f8c958cc2..c6fcf6d76951 100644 --- a/mm/secretmem.c +++ b/mm/secretmem.c @@ -5,6 +5,7 @@ #include #include #include +#include #include #include #include @@ -23,24 +24,66 @@ #define SECRETMEM_UNCACHED 0x2 =20 struct secretmem_ctx { + struct gen_pool *pool; unsigned int mode; }; =20 -static struct page *secretmem_alloc_page(gfp_t gfp) +static int secretmem_pool_increase(struct secretmem_ctx *ctx, gfp_t gfp) { - /* - * FIXME: use a cache of large pages to reduce the direct map - * fragmentation - */ - return alloc_page(gfp); + unsigned long nr_pages =3D (1 << HPAGE_PMD_ORDER); + struct gen_pool *pool =3D ctx->pool; + unsigned long addr; + struct page *page; + int err; + + page =3D alloc_pages(gfp, HPAGE_PMD_ORDER); + if (!page) + return -ENOMEM; + + addr =3D (unsigned long)page_address(page); + split_page(page, HPAGE_PMD_ORDER); + + err =3D gen_pool_add(pool, addr, HPAGE_PMD_SIZE, NUMA_NO_NODE); + if (err) { + __free_pages(page, HPAGE_PMD_ORDER); + return err; + } + + __kernel_map_pages(page, nr_pages, 0); + + return 0; +} + +static struct page *secretmem_alloc_page(struct secretmem_ctx *ctx, + gfp_t gfp) +{ + struct gen_pool *pool =3D ctx->pool; + unsigned long addr; + struct page *page; + int err; + + if (gen_pool_avail(pool) < PAGE_SIZE) { + err =3D secretmem_pool_increase(ctx, gfp); + if (err) + return NULL; + } + + addr =3D gen_pool_alloc(pool, PAGE_SIZE); + if (!addr) + return NULL; + + page =3D virt_to_page(addr); + get_page(page); + + return page; } =20 static vm_fault_t secretmem_fault(struct vm_fault *vmf) { + struct secretmem_ctx *ctx =3D vmf->vma->vm_file->private_data; struct address_space *mapping =3D vmf->vma->vm_file->f_mapping; struct inode *inode =3D file_inode(vmf->vma->vm_file); pgoff_t offset =3D vmf->pgoff; - unsigned long addr; struct page *page; int ret =3D 0; =20 @@ -49,7 +92,7 @@ static vm_fault_t secretmem_fault(struct vm_fault *vmf) =20 page =3D find_get_entry(mapping, offset); if (!page) { - page =3D secretmem_alloc_page(vmf->gfp_mask); + page =3D secretmem_alloc_page(ctx, vmf->gfp_mask); if (!page) return vmf_error(-ENOMEM); =20 @@ -57,14 +100,8 @@ static vm_fault_t secretmem_fault(struct vm_fault *vm= f) if (unlikely(ret)) goto err_put_page; =20 - ret =3D set_direct_map_invalid_noflush(page); - if (ret) - goto err_del_page_cache; - - addr =3D (unsigned long)page_address(page); - flush_tlb_kernel_range(addr, addr + PAGE_SIZE); - __SetPageUptodate(page); + set_page_private(page, (unsigned long)ctx); =20 ret =3D VM_FAULT_LOCKED; } @@ -72,8 +109,6 @@ static vm_fault_t secretmem_fault(struct vm_fault *vmf= ) vmf->page =3D page; return ret; =20 -err_del_page_cache: - delete_from_page_cache(page); err_put_page: put_page(page); return vmf_error(ret); @@ -155,7 +190,11 @@ static int secretmem_migratepage(struct address_spac= e *mapping, =20 static void secretmem_freepage(struct page *page) { - set_direct_map_default_noflush(page); + unsigned long addr =3D (unsigned long)page_address(page); + struct secretmem_ctx *ctx =3D (struct secretmem_ctx *)page_private(page= ); + struct gen_pool *pool =3D ctx->pool; + + gen_pool_free(pool, addr, PAGE_SIZE); } =20 static const struct address_space_operations secretmem_aops =3D { @@ -179,13 +218,18 @@ struct file *secretmem_file_create(const char *name= , unsigned int flags) if (!ctx) goto err_free_inode; =20 + ctx->pool =3D gen_pool_create(PAGE_SHIFT, NUMA_NO_NODE); + if (!ctx->pool) + goto err_free_ctx; + file =3D alloc_file_pseudo(inode, secretmem_mnt, "secretmem", O_RDWR, &secretmem_fops); if (IS_ERR(file)) - goto err_free_ctx; + goto err_free_pool; =20 mapping_set_unevictable(inode->i_mapping); =20 + inode->i_private =3D ctx; inode->i_mapping->private_data =3D ctx; inode->i_mapping->a_ops =3D &secretmem_aops; =20 @@ -197,6 +241,8 @@ struct file *secretmem_file_create(const char *name, = unsigned int flags) =20 return file; =20 +err_free_pool: + gen_pool_destroy(ctx->pool); err_free_ctx: kfree(ctx); err_free_inode: @@ -204,11 +250,34 @@ struct file *secretmem_file_create(const char *name= , unsigned int flags) return file; } =20 +static void secretmem_cleanup_chunk(struct gen_pool *pool, + struct gen_pool_chunk *chunk, void *data) +{ + unsigned long start =3D chunk->start_addr; + unsigned long end =3D chunk->end_addr; + unsigned long nr_pages, addr; + + nr_pages =3D (end - start + 1) / PAGE_SIZE; + __kernel_map_pages(virt_to_page(start), nr_pages, 1); + + for (addr =3D start; addr < end; addr +=3D PAGE_SIZE) + put_page(virt_to_page(addr)); +} + +static void secretmem_cleanup_pool(struct secretmem_ctx *ctx) +{ + struct gen_pool *pool =3D ctx->pool; + + gen_pool_for_each_chunk(pool, secretmem_cleanup_chunk, ctx); + gen_pool_destroy(pool); +} + static void secretmem_evict_inode(struct inode *inode) { struct secretmem_ctx *ctx =3D inode->i_private; =20 truncate_inode_pages_final(&inode->i_data); + secretmem_cleanup_pool(ctx); clear_inode(inode); kfree(ctx); } --=20 2.26.2