Received: by 2002:a05:6a10:8c0a:0:0:0:0 with SMTP id go10csp2772185pxb; Tue, 9 Mar 2021 10:24:45 -0800 (PST) X-Google-Smtp-Source: ABdhPJzMSnfeSlt2h1ofYwE4j0dszuxOciQ8aCsQGjvy6eckJo79SgMtnvRx2qcEbgPB2HhdCu1Z X-Received: by 2002:a50:fb10:: with SMTP id d16mr5713074edq.73.1615314285649; Tue, 09 Mar 2021 10:24:45 -0800 (PST) ARC-Seal: i=1; a=rsa-sha256; t=1615314285; cv=none; d=google.com; s=arc-20160816; b=gllLDrTDK+Kz7FUDi8Zz3vSa3xUP0euoFztA71FegCkSEa2c7ZBt7sFw2oqnFRE/KQ E4KLFBq7HCanT9pMYOG5WqEOGoNzpBVrnxCZFiP26hLmlsziAXS1Pi6hbRQeIa4fWTz7 heX0/ZMNofB7kcKQnid1calOSZrGUj+jZVpFFVIcPiqAfa8caU1lFH/mQEXXSxUhE480 gm7IJ9SLPGP41FYDYQg5Sk0LKejcQ2HBqyk6wOItAI29In+oqqowKTk5ZdG1TWUA4tSE 0cQyqlevh710LvWrSrT7D6Sl3TP6e3Z2adpUt1MO8ciTMDUiW7IGeYxsXKJUOPET+efO taeA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:content-transfer-encoding:mime-version :references:in-reply-to:message-id:date:subject:cc:to:from :dkim-signature; bh=04qMOwDZliJ4nJyv6wec/a7UlSmhF51F/NyAZPQ4Q9w=; b=K6mg3T76qHe74I/iv/dRbM2dTiuNGnP54umxteLkGom/Gk0VzIYMdxqRUtNiudFKO4 HaUO1KExceV8U0sKk17LSlzntg+iBxdces70JZLa+sc+LwjF0KBarT45HYrCDMLAKNNC 4h8EWNunp2bOBIIFOwMD8loCdsX1V/b/LEukGN4UxT0JxRWerFh/LXMg3k9bZSc9Lsbr D2pi29CZuGlE79bog81HwPh5wqxnWlY7f0DTPAWUrzWmhDvxgFi6PCOZxjmyiVLEYePe a0XBx6CfV5jYcQwKz5pou5WvFX5K4rMKFh5DhdCNIgY5hmaAFBy5Fp1GlBTIooO+qme4 sPyQ== ARC-Authentication-Results: i=1; mx.google.com; dkim=pass header.i=@yandex-team.com header.s=default header.b="y3dHOgf/"; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=yandex-team.com Return-Path: Received: from vger.kernel.org (vger.kernel.org. [23.128.96.18]) by mx.google.com with ESMTP id eb13si10204697edb.315.2021.03.09.10.24.23; Tue, 09 Mar 2021 10:24:45 -0800 (PST) Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18; Authentication-Results: mx.google.com; dkim=pass header.i=@yandex-team.com header.s=default header.b="y3dHOgf/"; spf=pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org; dmarc=pass (p=NONE sp=NONE dis=NONE) header.from=yandex-team.com Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230451AbhCISX3 (ORCPT + 99 others); Tue, 9 Mar 2021 13:23:29 -0500 Received: from forwardcorp1p.mail.yandex.net ([77.88.29.217]:46696 "EHLO forwardcorp1p.mail.yandex.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230403AbhCISXK (ORCPT ); Tue, 9 Mar 2021 13:23:10 -0500 Received: from iva8-d077482f1536.qloud-c.yandex.net (iva8-d077482f1536.qloud-c.yandex.net [IPv6:2a02:6b8:c0c:2f26:0:640:d077:482f]) by forwardcorp1p.mail.yandex.net (Yandex) with ESMTP id 7FB332E150B; Tue, 9 Mar 2021 21:23:05 +0300 (MSK) Received: from iva4-f06c35e68a0a.qloud-c.yandex.net (iva4-f06c35e68a0a.qloud-c.yandex.net [2a02:6b8:c0c:152e:0:640:f06c:35e6]) by iva8-d077482f1536.qloud-c.yandex.net (mxbackcorp/Yandex) with ESMTP id wcnaa8RXHT-N5JiONba; Tue, 09 Mar 2021 21:23:05 +0300 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=yandex-team.com; s=default; t=1615314185; bh=04qMOwDZliJ4nJyv6wec/a7UlSmhF51F/NyAZPQ4Q9w=; h=In-Reply-To:Message-Id:References:Date:Subject:To:From:Cc; b=y3dHOgf/5hmNhvNu/3cGc5E2ES5yO94+MLGeeSxzMHmf3AkeVwEpV7RSDc1d/jBJB bpDqXXsW19wWkvUqO5c3BEw2tXmHnpkKjDLT0EcZ+hs9TZudU3nSCcRxVNcXwnyHm1 8JtkWgXIL6YFaF7CGq6yLvtueQ/NKSuaad+YAHQo= Authentication-Results: iva8-d077482f1536.qloud-c.yandex.net; dkim=pass header.i=@yandex-team.com Received: from dynamic-vpn.dhcp.yndx.net (dynamic-vpn.dhcp.yndx.net [2a02:6b8:b081:203::1:15]) by iva4-f06c35e68a0a.qloud-c.yandex.net (smtpcorp/Yandex) with ESMTPSA id T5bJwzP5HK-N4nWQO9P; Tue, 09 Mar 2021 21:23:04 +0300 (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)) (Client certificate not present) From: Andrey Ryabinin To: gregkh@linuxfoundation.org, stable@vger.kernel.org Cc: jroedel@suse.de, will@kernel.org, linux-kernel@vger.kernel.org, Andrey Ryabinin Subject: [PATCH stable 4.14-4.19] iommu/amd: Fix sleeping in atomic in increase_address_space() Date: Tue, 9 Mar 2021 21:24:27 +0300 Message-Id: <20210309182430.18849-1-arbn@yandex-team.com> X-Mailer: git-send-email 2.26.2 In-Reply-To: <161512522533161@kroah.com> References: <161512522533161@kroah.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org commit 140456f994195b568ecd7fc2287a34eadffef3ca upstream. increase_address_space() calls get_zeroed_page(gfp) under spin_lock with disabled interrupts. gfp flags passed to increase_address_space() may allow sleeping, so it comes to this: BUG: sleeping function called from invalid context at mm/page_alloc.c:4342 in_atomic(): 1, irqs_disabled(): 1, pid: 21555, name: epdcbbf1qnhbsd8 Call Trace: dump_stack+0x66/0x8b ___might_sleep+0xec/0x110 __alloc_pages_nodemask+0x104/0x300 get_zeroed_page+0x15/0x40 iommu_map_page+0xdd/0x3e0 amd_iommu_map+0x50/0x70 iommu_map+0x106/0x220 vfio_iommu_type1_ioctl+0x76e/0x950 [vfio_iommu_type1] do_vfs_ioctl+0xa3/0x6f0 ksys_ioctl+0x66/0x70 __x64_sys_ioctl+0x16/0x20 do_syscall_64+0x4e/0x100 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Fix this by moving get_zeroed_page() out of spin_lock/unlock section. Fixes: 754265bcab ("iommu/amd: Fix race in increase_address_space()") Signed-off-by: Andrey Ryabinin Acked-by: Will Deacon Link: https://lore.kernel.org/r/20210217143004.19165-1-arbn@yandex-team.com Signed-off-by: Joerg Roedel Signed-off-by: Andrey Ryabinin --- drivers/iommu/amd_iommu.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c index 494caaa265af0..8195ff219b48c 100644 --- a/drivers/iommu/amd_iommu.c +++ b/drivers/iommu/amd_iommu.c @@ -1347,24 +1347,26 @@ static void increase_address_space(struct protection_domain *domain, unsigned long flags; u64 *pte; + pte = (void *)get_zeroed_page(gfp); + if (!pte) + return; + spin_lock_irqsave(&domain->lock, flags); if (WARN_ON_ONCE(domain->mode == PAGE_MODE_6_LEVEL)) /* address space already 64 bit large */ goto out; - pte = (void *)get_zeroed_page(gfp); - if (!pte) - goto out; - *pte = PM_LEVEL_PDE(domain->mode, iommu_virt_to_phys(domain->pt_root)); domain->pt_root = pte; domain->mode += 1; domain->updated = true; + pte = NULL; out: spin_unlock_irqrestore(&domain->lock, flags); + free_page((unsigned long)pte); return; } -- 2.26.2