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 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 49714C433F5 for ; Thu, 28 Oct 2021 18:41:04 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id B34DE60D43 for ; Thu, 28 Oct 2021 18:41:03 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.4.1 mail.kernel.org B34DE60D43 Authentication-Results: mail.kernel.org; dmarc=fail (p=reject dis=none) header.from=google.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=kvack.org Received: by kanga.kvack.org (Postfix) id DFF786B0071; Thu, 28 Oct 2021 14:41:02 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id DAED4940007; Thu, 28 Oct 2021 14:41:02 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id C76866B0073; Thu, 28 Oct 2021 14:41:02 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0108.hostedemail.com [216.40.44.108]) by kanga.kvack.org (Postfix) with ESMTP id A0C7F6B0071 for ; Thu, 28 Oct 2021 14:41:02 -0400 (EDT) Received: from smtpin21.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay03.hostedemail.com (Postfix) with ESMTP id 276588249980 for ; Thu, 28 Oct 2021 18:41:02 +0000 (UTC) X-FDA: 78746713164.21.2B44D09 Received: from mail-yb1-f178.google.com (mail-yb1-f178.google.com [209.85.219.178]) by imf22.hostedemail.com (Postfix) with ESMTP id BFD79192B for ; Thu, 28 Oct 2021 18:41:01 +0000 (UTC) Received: by mail-yb1-f178.google.com with SMTP id v64so11231249ybi.5 for ; Thu, 28 Oct 2021 11:41:01 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20210112; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=U6hI4QUgyE2eFqGR1tRgs9ayNRlgsskzXHlchXAdB2g=; b=lZWgvDCv1CJD+W7u2YU/u9zSaI1ddrKl8GY1ChLNEQCV3IT7cYI+bPKS6yz5lJi0De R1iH0eAwjKmN99YnNfLK5JF1i51IKFGtdytLKl/ggXr+82O7kUiG0cKdtGXOQpbT57Yj skMa2Sli6rTHq0jFfdZO2BXN543N6lOY1fqI+4KxSSuFFuFqg/T16rGiQnFQkXAdZ5mF Cd1/D2iVSSoQKT1t9/0TeY+auqQvFFWQd80pT3fAEW1T+C15CE/3N9dPlSecLnp3Lezj YLAFaHEEARpBE5dZAGMpxnexQ7jwKTBJz+ZGLIjp5ZihScuBc1/kPzWze+mX3s1RZt6E ptBg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=U6hI4QUgyE2eFqGR1tRgs9ayNRlgsskzXHlchXAdB2g=; b=wm3GbFzeF7WtiNQ0U40oCIYC+6/vyh6VSXAMvDziePfuS5oOP+MPDFPuP69cM9avew ZrBaUcAXcmzB/gL02/CuN3o6/MWpRHTU8HcgbMMU0TxDnk3tQmcEMKZgE3XenJHr0oPH 5S+fMpf9Q3L5G6n6jZayQd28V6HAEEAjYcGWEWVGvbVHXdw1QlwLZKr83Fe/ghdyV4yd +nXrCBakx8Wc13XganfJZYU5EdkQTPH3DihPxRXs8STaZR3Qz5cNFZ9iTobYe/t1ZRcs rA5e8I0bHkpzMDXYkmhj71GacMan8qcIM2VbWfEIyC4yX+2kk7Ff7teF7yzn/Tsv1Woc ntUA== X-Gm-Message-State: AOAM532VpTcT++Nf4MoAGW2mMGm4wfaxam54edfRr1l4m7xoyxyp80Rk vuCHoWui5AJZ9o27ZV2NmTKXKmMN2LTu3IQVqX2HUQ== X-Google-Smtp-Source: ABdhPJwhle1+VYAuOAeoY+R8+v8PMmi3ITn2fM/eF7pVouM03UKR4EPnPRSw6t9ZP2APlrrrqpQLi5XUuIHU2uOysoc= X-Received: by 2002:a25:c5c5:: with SMTP id v188mr6417772ybe.34.1635446460719; Thu, 28 Oct 2021 11:41:00 -0700 (PDT) MIME-Version: 1.0 References: <1633701982-22302-1-git-send-email-charante@codeaurora.org> In-Reply-To: <1633701982-22302-1-git-send-email-charante@codeaurora.org> From: Suren Baghdasaryan Date: Thu, 28 Oct 2021 11:40:49 -0700 Message-ID: Subject: Re: [PATCH] mm: shmem: implement POSIX_FADV_[WILL|DONT]NEED for shmem To: Charan Teja Reddy Cc: hughd@google.com, akpm@linux-foundation.org, vbabka@suse.cz, rientjes@google.com, david@redhat.com, mhocko@suse.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org Content-Type: text/plain; charset="UTF-8" X-Rspamd-Server: rspam05 X-Rspamd-Queue-Id: BFD79192B X-Stat-Signature: 5879ycguf91x8arc3uwktpwk3oredjh8 Authentication-Results: imf22.hostedemail.com; dkim=pass header.d=google.com header.s=20210112 header.b=lZWgvDCv; spf=pass (imf22.hostedemail.com: domain of surenb@google.com designates 209.85.219.178 as permitted sender) smtp.mailfrom=surenb@google.com; dmarc=pass (policy=reject) header.from=google.com X-HE-Tag: 1635446461-46235 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: On Fri, Oct 8, 2021 at 7:07 AM Charan Teja Reddy wrote: > > Currently fadvise(2) is supported only for the files that doesn't > associated with noop_backing_dev_info thus for the files, like shmem, > fadvise results into NOP. But then there is file_operations->fadvise() > that lets the file systems to implement their own fadvise > implementation. Use this support to implement some of the POSIX_FADV_XXX > functionality for shmem files. > > This patch aims to implement POSIX_FADV_WILLNEED and POSIX_FADV_DONTNEED > advices to shmem files which can be helpful for the drivers who may want > to manage the shmem pages of the files that are created through > shmem_file_setup[_with_mnt](). An example usecase may be like, driver > can create the shmem file of the size equal to its requirements and > map the pages for DMA and then pass the fd to user. The user who knows > well about the usage of these pages can now decide when these pages are > not required push them to swap through DONTNEED thus free up memory well > in advance rather than relying on the reclaim and use WILLNEED when it > decide that they are useful in the near future. IOW, it lets the clients > to free up/read the memory when it wants to. Another usecase is that GEM > objets which are currenlty allocated and managed through shmem files can > use vfs_fadvise(DONT|WILLNEED) on shmem fd when the driver comes to > know(like through some hints from user space) that GEM objects are not > going to use/will need in the near future. Sounds useful. You should probably add a documentation section for manual pages. > > Signed-off-by: Charan Teja Reddy > --- > mm/shmem.c | 139 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > 1 file changed, 139 insertions(+) > > diff --git a/mm/shmem.c b/mm/shmem.c > index 70d9ce2..ab7ea33 100644 > --- a/mm/shmem.c > +++ b/mm/shmem.c > @@ -38,6 +38,8 @@ > #include > #include > #include > +#include > +#include > > #include /* for arch/microblaze update_mmu_cache() */ > > @@ -2792,6 +2794,142 @@ static long shmem_fallocate(struct file *file, int mode, loff_t offset, > return error; > } > > +static int shmem_fadvise_dontneed(struct address_space *mapping, loff_t start, > + loff_t end) > +{ > + int ret; > + struct writeback_control wbc = { > + .sync_mode = WB_SYNC_NONE, > + .nr_to_write = LONG_MAX, > + .range_start = 0, > + .range_end = LLONG_MAX, > + .for_reclaim = 1, > + }; > + struct page *page; > + > + XA_STATE(xas, &mapping->i_pages, start); XA_STATE is a variable definition, so I think the empty line should be moved down here. > + if (!shmem_mapping(mapping)) > + return -EINVAL; > + > + if (!total_swap_pages) > + return 0; > + > + lru_add_drain(); > + > + rcu_read_lock(); > + xas_for_each(&xas, page, end) { > + if (xas_retry(&xas, page)) > + continue; > + if (xa_is_value(page)) > + continue; > + if (isolate_lru_page(page)) > + continue; isolate_lru_page comment says that it "Must be called with an elevated refcount on the page". Are you ensuring that? Also not sure if you need to isolate the page before this writeback. > + > + inc_node_page_state(page, NR_ISOLATED_ANON + > + page_is_file_lru(page)); > + lock_page(page); > + ClearPageDirty(page); I see that other places use clear_page_dirty_for_io() to clear the dirty flag before writepage(), don't you need it here? > + SetPageReclaim(page); > + ret = shmem_writepage(page, &wbc); > + if (!PageWriteback(page)) > + ClearPageReclaim(page); > + if (ret) { > + unlock_page(page); > + putback_lru_page(page); > + dec_node_page_state(page, NR_ISOLATED_ANON + > + page_is_file_lru(page)); > + continue; > + } > + > + /* > + * shmem_writepage() place the page in the swapcache. > + * Delete the page from the swapcache and release the > + * page. Won't deleting the page from swap cache interfere with ongoing writeback if it has not yet completed? > + */ > + lock_page(page); You have the page already locked, the above seems extra. > + delete_from_swap_cache(page); > + unlock_page(page); > + dec_node_page_state(page, NR_ISOLATED_ANON + > + page_is_file_lru(page)); > + put_page(page); > + if (need_resched()) { > + xas_pause(&xas); > + cond_resched_rcu(); > + } > + } > + rcu_read_unlock(); > + > + return 0; > +} > + > +static int shmem_fadvise_willneed(struct address_space *mapping, > + pgoff_t start, pgoff_t long end) > +{ > + struct page *page; > + > + XA_STATE(xas, &mapping->i_pages, start); Empty line should be moved here. > + rcu_read_lock(); > + xas_for_each(&xas, page, end) { > + if (!xa_is_value(page)) > + continue; > + page = shmem_read_mapping_page(mapping, xas.xa_index); > + if (!IS_ERR(page)) > + put_page(page); > + > + if (need_resched()) { > + xas_pause(&xas); > + cond_resched_rcu(); > + } > + } > + rcu_read_unlock(); > + > + return 0; > +} > + > +static int shmem_fadvise(struct file *file, loff_t offset, loff_t len, int advice) > +{ > + loff_t endbyte; > + pgoff_t start_index; > + pgoff_t end_index; > + struct address_space *mapping; > + int ret = 0; > + > + mapping = file->f_mapping; > + if (!mapping || len < 0) > + return -EINVAL; > + > + endbyte = (u64)offset + (u64)len; Why are these typecasts needed? All involved vars seem to be of the same loff_t type. Are you worried about overflow after addition? If so why not just: if (offset + len < 0) return -EINVAL; > + if (!len || endbyte < len) > + endbyte = -1; > + else > + endbyte--; You already checked that len is not negative. You could simply add: if (!len) return 0; > + > + > + start_index = offset >> PAGE_SHIFT; > + end_index = endbyte >> PAGE_SHIFT; > + switch (advice) { > + case POSIX_FADV_DONTNEED: > + ret = shmem_fadvise_dontneed(mapping, start_index, end_index); > + break; > + case POSIX_FADV_WILLNEED: > + ret = shmem_fadvise_willneed(mapping, start_index, end_index); > + break; > + case POSIX_FADV_NORMAL: > + case POSIX_FADV_RANDOM: > + case POSIX_FADV_SEQUENTIAL: > + case POSIX_FADV_NOREUSE: > + /* > + * No bad return value, but ignore advice. May have to > + * implement in future. > + */ Returning ENOSYS might be better here IMHO. When some of these options are implemented, the userspace will be able to detect if the kernel supports it by checking for that error code. If you return 0 in both cases such detection will be impossible. > + break; > + default: > + return -EINVAL; > + } > + > + return ret; > +} > + > static int shmem_statfs(struct dentry *dentry, struct kstatfs *buf) > { > struct shmem_sb_info *sbinfo = SHMEM_SB(dentry->d_sb); > @@ -3799,6 +3937,7 @@ static const struct file_operations shmem_file_operations = { > .splice_write = iter_file_splice_write, > .fallocate = shmem_fallocate, > #endif > + .fadvise = shmem_fadvise, > }; > > static const struct inode_operations shmem_inode_operations = { > -- > QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a > member of the Code Aurora Forum, hosted by The Linux Foundation > >