From patchwork Wed Jul 1 12:16:37 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: =?utf-8?b?SsO8cmdlbiBHcm/Dnw==?= X-Patchwork-Id: 11636213 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 2CAC213B4 for ; Wed, 1 Jul 2020 12:23:47 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 11CC9206CB for ; Wed, 1 Jul 2020 12:23:47 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 11CC9206CB Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=suse.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1jqbfQ-00058r-NO; Wed, 01 Jul 2020 12:16:44 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1jqbfP-00058k-Kn for xen-devel@lists.xenproject.org; Wed, 01 Jul 2020 12:16:43 +0000 X-Inumbo-ID: b8f40f48-bb94-11ea-8496-bc764e2007e4 Received: from mx2.suse.de (unknown [195.135.220.15]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id b8f40f48-bb94-11ea-8496-bc764e2007e4; Wed, 01 Jul 2020 12:16:42 +0000 (UTC) X-Virus-Scanned: by amavisd-new at test-mx.suse.de Received: from relay2.suse.de (unknown [195.135.221.27]) by mx2.suse.de (Postfix) with ESMTP id CA5A5ADE1; Wed, 1 Jul 2020 12:16:41 +0000 (UTC) From: Juergen Gross To: xen-devel@lists.xenproject.org, linux-kernel@vger.kernel.org, clang-built-linux@googlegroups.com Subject: [PATCH v2 1/2] xen/xenbus: avoid large structs and arrays on the stack Date: Wed, 1 Jul 2020 14:16:37 +0200 Message-Id: <20200701121638.19840-2-jgross@suse.com> X-Mailer: git-send-email 2.26.2 In-Reply-To: <20200701121638.19840-1-jgross@suse.com> References: <20200701121638.19840-1-jgross@suse.com> MIME-Version: 1.0 X-BeenThere: xen-devel@lists.xenproject.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Cc: Juergen Gross , Boris Ostrovsky , Stefano Stabellini , Arnd Bergmann Errors-To: xen-devel-bounces@lists.xenproject.org Sender: "Xen-devel" xenbus_map_ring_valloc() and its sub-functions are putting quite large structs and arrays on the stack. This is problematic at runtime, but might also result in build failures (e.g. with clang due to the option -Werror,-Wframe-larger-than=... used). Fix that by moving most of the data from the stack into a dynamically allocated struct. Performance is no issue here, as xenbus_map_ring_valloc() is used only when adding a new PV device to a backend driver. While at it move some duplicated code from pv/hvm specific mapping functions to the single caller. Reported-by: Arnd Bergmann Signed-off-by: Juergen Gross Reviewed-by: Boris Ostrovsky --- V2: - shorten internal function names (Boris Ostrovsky) --- drivers/xen/xenbus/xenbus_client.c | 161 +++++++++++++++-------------- 1 file changed, 83 insertions(+), 78 deletions(-) diff --git a/drivers/xen/xenbus/xenbus_client.c b/drivers/xen/xenbus/xenbus_client.c index 040d2a43e8e3..9f8372079ecf 100644 --- a/drivers/xen/xenbus/xenbus_client.c +++ b/drivers/xen/xenbus/xenbus_client.c @@ -69,11 +69,27 @@ struct xenbus_map_node { unsigned int nr_handles; }; +struct map_ring_valloc { + struct xenbus_map_node *node; + + /* Why do we need two arrays? See comment of __xenbus_map_ring */ + union { + unsigned long addrs[XENBUS_MAX_RING_GRANTS]; + pte_t *ptes[XENBUS_MAX_RING_GRANTS]; + }; + phys_addr_t phys_addrs[XENBUS_MAX_RING_GRANTS]; + + struct gnttab_map_grant_ref map[XENBUS_MAX_RING_GRANTS]; + struct gnttab_unmap_grant_ref unmap[XENBUS_MAX_RING_GRANTS]; + + unsigned int idx; /* HVM only. */ +}; + static DEFINE_SPINLOCK(xenbus_valloc_lock); static LIST_HEAD(xenbus_valloc_pages); struct xenbus_ring_ops { - int (*map)(struct xenbus_device *dev, + int (*map)(struct xenbus_device *dev, struct map_ring_valloc *info, grant_ref_t *gnt_refs, unsigned int nr_grefs, void **vaddr); int (*unmap)(struct xenbus_device *dev, void *vaddr); @@ -449,12 +465,32 @@ int xenbus_map_ring_valloc(struct xenbus_device *dev, grant_ref_t *gnt_refs, unsigned int nr_grefs, void **vaddr) { int err; + struct map_ring_valloc *info; + + *vaddr = NULL; + + if (nr_grefs > XENBUS_MAX_RING_GRANTS) + return -EINVAL; + + info = kzalloc(sizeof(*info), GFP_KERNEL); + if (!info) + return -ENOMEM; + + info->node = kzalloc(sizeof(*info->node), GFP_KERNEL); + if (!info->node) { + err = -ENOMEM; + goto out; + } + + err = ring_ops->map(dev, info, gnt_refs, nr_grefs, vaddr); - err = ring_ops->map(dev, gnt_refs, nr_grefs, vaddr); /* Some hypervisors are buggy and can return 1. */ if (err > 0) err = GNTST_general_error; + out: + kfree(info->node); + kfree(info); return err; } EXPORT_SYMBOL_GPL(xenbus_map_ring_valloc); @@ -466,12 +502,10 @@ static int __xenbus_map_ring(struct xenbus_device *dev, grant_ref_t *gnt_refs, unsigned int nr_grefs, grant_handle_t *handles, - phys_addr_t *addrs, + struct map_ring_valloc *info, unsigned int flags, bool *leaked) { - struct gnttab_map_grant_ref map[XENBUS_MAX_RING_GRANTS]; - struct gnttab_unmap_grant_ref unmap[XENBUS_MAX_RING_GRANTS]; int i, j; int err = GNTST_okay; @@ -479,23 +513,22 @@ static int __xenbus_map_ring(struct xenbus_device *dev, return -EINVAL; for (i = 0; i < nr_grefs; i++) { - memset(&map[i], 0, sizeof(map[i])); - gnttab_set_map_op(&map[i], addrs[i], flags, gnt_refs[i], - dev->otherend_id); + gnttab_set_map_op(&info->map[i], info->phys_addrs[i], flags, + gnt_refs[i], dev->otherend_id); handles[i] = INVALID_GRANT_HANDLE; } - gnttab_batch_map(map, i); + gnttab_batch_map(info->map, i); for (i = 0; i < nr_grefs; i++) { - if (map[i].status != GNTST_okay) { - err = map[i].status; - xenbus_dev_fatal(dev, map[i].status, + if (info->map[i].status != GNTST_okay) { + err = info->map[i].status; + xenbus_dev_fatal(dev, info->map[i].status, "mapping in shared page %d from domain %d", gnt_refs[i], dev->otherend_id); goto fail; } else - handles[i] = map[i].handle; + handles[i] = info->map[i].handle; } return GNTST_okay; @@ -503,19 +536,19 @@ static int __xenbus_map_ring(struct xenbus_device *dev, fail: for (i = j = 0; i < nr_grefs; i++) { if (handles[i] != INVALID_GRANT_HANDLE) { - memset(&unmap[j], 0, sizeof(unmap[j])); - gnttab_set_unmap_op(&unmap[j], (phys_addr_t)addrs[i], + gnttab_set_unmap_op(&info->unmap[j], + info->phys_addrs[i], GNTMAP_host_map, handles[i]); j++; } } - if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, unmap, j)) + if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, info->unmap, j)) BUG(); *leaked = false; for (i = 0; i < j; i++) { - if (unmap[i].status != GNTST_okay) { + if (info->unmap[i].status != GNTST_okay) { *leaked = true; break; } @@ -566,21 +599,12 @@ static int xenbus_unmap_ring(struct xenbus_device *dev, grant_handle_t *handles, return err; } -struct map_ring_valloc_hvm -{ - unsigned int idx; - - /* Why do we need two arrays? See comment of __xenbus_map_ring */ - phys_addr_t phys_addrs[XENBUS_MAX_RING_GRANTS]; - unsigned long addrs[XENBUS_MAX_RING_GRANTS]; -}; - static void xenbus_map_ring_setup_grant_hvm(unsigned long gfn, unsigned int goffset, unsigned int len, void *data) { - struct map_ring_valloc_hvm *info = data; + struct map_ring_valloc *info = data; unsigned long vaddr = (unsigned long)gfn_to_virt(gfn); info->phys_addrs[info->idx] = vaddr; @@ -589,39 +613,28 @@ static void xenbus_map_ring_setup_grant_hvm(unsigned long gfn, info->idx++; } -static int xenbus_map_ring_valloc_hvm(struct xenbus_device *dev, - grant_ref_t *gnt_ref, - unsigned int nr_grefs, - void **vaddr) +static int xenbus_map_ring_hvm(struct xenbus_device *dev, + struct map_ring_valloc *info, + grant_ref_t *gnt_ref, + unsigned int nr_grefs, + void **vaddr) { - struct xenbus_map_node *node; + struct xenbus_map_node *node = info->node; int err; void *addr; bool leaked = false; - struct map_ring_valloc_hvm info = { - .idx = 0, - }; unsigned int nr_pages = XENBUS_PAGES(nr_grefs); - if (nr_grefs > XENBUS_MAX_RING_GRANTS) - return -EINVAL; - - *vaddr = NULL; - - node = kzalloc(sizeof(*node), GFP_KERNEL); - if (!node) - return -ENOMEM; - err = alloc_xenballooned_pages(nr_pages, node->hvm.pages); if (err) goto out_err; gnttab_foreach_grant(node->hvm.pages, nr_grefs, xenbus_map_ring_setup_grant_hvm, - &info); + info); err = __xenbus_map_ring(dev, gnt_ref, nr_grefs, node->handles, - info.phys_addrs, GNTMAP_host_map, &leaked); + info, GNTMAP_host_map, &leaked); node->nr_handles = nr_grefs; if (err) @@ -641,11 +654,13 @@ static int xenbus_map_ring_valloc_hvm(struct xenbus_device *dev, spin_unlock(&xenbus_valloc_lock); *vaddr = addr; + info->node = NULL; + return 0; out_xenbus_unmap_ring: if (!leaked) - xenbus_unmap_ring(dev, node->handles, nr_grefs, info.addrs); + xenbus_unmap_ring(dev, node->handles, nr_grefs, info->addrs); else pr_alert("leaking %p size %u page(s)", addr, nr_pages); @@ -653,7 +668,6 @@ static int xenbus_map_ring_valloc_hvm(struct xenbus_device *dev, if (!leaked) free_xenballooned_pages(nr_pages, node->hvm.pages); out_err: - kfree(node); return err; } @@ -676,40 +690,30 @@ int xenbus_unmap_ring_vfree(struct xenbus_device *dev, void *vaddr) EXPORT_SYMBOL_GPL(xenbus_unmap_ring_vfree); #ifdef CONFIG_XEN_PV -static int xenbus_map_ring_valloc_pv(struct xenbus_device *dev, - grant_ref_t *gnt_refs, - unsigned int nr_grefs, - void **vaddr) +static int xenbus_map_ring_pv(struct xenbus_device *dev, + struct map_ring_valloc *info, + grant_ref_t *gnt_refs, + unsigned int nr_grefs, + void **vaddr) { - struct xenbus_map_node *node; + struct xenbus_map_node *node = info->node; struct vm_struct *area; - pte_t *ptes[XENBUS_MAX_RING_GRANTS]; - phys_addr_t phys_addrs[XENBUS_MAX_RING_GRANTS]; int err = GNTST_okay; int i; bool leaked; - *vaddr = NULL; - - if (nr_grefs > XENBUS_MAX_RING_GRANTS) - return -EINVAL; - - node = kzalloc(sizeof(*node), GFP_KERNEL); - if (!node) - return -ENOMEM; - - area = alloc_vm_area(XEN_PAGE_SIZE * nr_grefs, ptes); + area = alloc_vm_area(XEN_PAGE_SIZE * nr_grefs, info->ptes); if (!area) { kfree(node); return -ENOMEM; } for (i = 0; i < nr_grefs; i++) - phys_addrs[i] = arbitrary_virt_to_machine(ptes[i]).maddr; + info->phys_addrs[i] = + arbitrary_virt_to_machine(info->ptes[i]).maddr; err = __xenbus_map_ring(dev, gnt_refs, nr_grefs, node->handles, - phys_addrs, - GNTMAP_host_map | GNTMAP_contains_pte, + info, GNTMAP_host_map | GNTMAP_contains_pte, &leaked); if (err) goto failed; @@ -722,6 +726,8 @@ static int xenbus_map_ring_valloc_pv(struct xenbus_device *dev, spin_unlock(&xenbus_valloc_lock); *vaddr = area->addr; + info->node = NULL; + return 0; failed: @@ -730,11 +736,10 @@ static int xenbus_map_ring_valloc_pv(struct xenbus_device *dev, else pr_alert("leaking VM area %p size %u page(s)", area, nr_grefs); - kfree(node); return err; } -static int xenbus_unmap_ring_vfree_pv(struct xenbus_device *dev, void *vaddr) +static int xenbus_unmap_ring_pv(struct xenbus_device *dev, void *vaddr) { struct xenbus_map_node *node; struct gnttab_unmap_grant_ref unmap[XENBUS_MAX_RING_GRANTS]; @@ -798,12 +803,12 @@ static int xenbus_unmap_ring_vfree_pv(struct xenbus_device *dev, void *vaddr) } static const struct xenbus_ring_ops ring_ops_pv = { - .map = xenbus_map_ring_valloc_pv, - .unmap = xenbus_unmap_ring_vfree_pv, + .map = xenbus_map_ring_pv, + .unmap = xenbus_unmap_ring_pv, }; #endif -struct unmap_ring_vfree_hvm +struct unmap_ring_hvm { unsigned int idx; unsigned long addrs[XENBUS_MAX_RING_GRANTS]; @@ -814,19 +819,19 @@ static void xenbus_unmap_ring_setup_grant_hvm(unsigned long gfn, unsigned int len, void *data) { - struct unmap_ring_vfree_hvm *info = data; + struct unmap_ring_hvm *info = data; info->addrs[info->idx] = (unsigned long)gfn_to_virt(gfn); info->idx++; } -static int xenbus_unmap_ring_vfree_hvm(struct xenbus_device *dev, void *vaddr) +static int xenbus_unmap_ring_hvm(struct xenbus_device *dev, void *vaddr) { int rv; struct xenbus_map_node *node; void *addr; - struct unmap_ring_vfree_hvm info = { + struct unmap_ring_hvm info = { .idx = 0, }; unsigned int nr_pages; @@ -887,8 +892,8 @@ enum xenbus_state xenbus_read_driver_state(const char *path) EXPORT_SYMBOL_GPL(xenbus_read_driver_state); static const struct xenbus_ring_ops ring_ops_hvm = { - .map = xenbus_map_ring_valloc_hvm, - .unmap = xenbus_unmap_ring_vfree_hvm, + .map = xenbus_map_ring_hvm, + .unmap = xenbus_unmap_ring_hvm, }; void __init xenbus_ring_ops_init(void) From patchwork Wed Jul 1 12:16:38 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: =?utf-8?b?SsO8cmdlbiBHcm/Dnw==?= X-Patchwork-Id: 11636209 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 6B55F913 for ; Wed, 1 Jul 2020 12:17:45 +0000 (UTC) Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 5159A20760 for ; Wed, 1 Jul 2020 12:17:45 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 5159A20760 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=suse.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=xen-devel-bounces@lists.xenproject.org Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1jqbfV-00059M-V7; Wed, 01 Jul 2020 12:16:49 +0000 Received: from us1-rack-iad1.inumbo.com ([172.99.69.81]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1jqbfU-00058k-JN for xen-devel@lists.xenproject.org; Wed, 01 Jul 2020 12:16:48 +0000 X-Inumbo-ID: b95029ae-bb94-11ea-bb8b-bc764e2007e4 Received: from mx2.suse.de (unknown [195.135.220.15]) by us1-rack-iad1.inumbo.com (Halon) with ESMTPS id b95029ae-bb94-11ea-bb8b-bc764e2007e4; Wed, 01 Jul 2020 12:16:43 +0000 (UTC) X-Virus-Scanned: by amavisd-new at test-mx.suse.de Received: from relay2.suse.de (unknown [195.135.221.27]) by mx2.suse.de (Postfix) with ESMTP id 6FE32ADE2; Wed, 1 Jul 2020 12:16:42 +0000 (UTC) From: Juergen Gross To: xen-devel@lists.xenproject.org, linux-kernel@vger.kernel.org Subject: [PATCH v2 2/2] xen/xenbus: let xenbus_map_ring_valloc() return errno values only Date: Wed, 1 Jul 2020 14:16:38 +0200 Message-Id: <20200701121638.19840-3-jgross@suse.com> X-Mailer: git-send-email 2.26.2 In-Reply-To: <20200701121638.19840-1-jgross@suse.com> References: <20200701121638.19840-1-jgross@suse.com> MIME-Version: 1.0 X-BeenThere: xen-devel@lists.xenproject.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Cc: Juergen Gross , Boris Ostrovsky , Stefano Stabellini Errors-To: xen-devel-bounces@lists.xenproject.org Sender: "Xen-devel" Today xenbus_map_ring_valloc() can return either a negative errno value (-ENOMEM or -EINVAL) or a grant status value. This is a mess as e.g -ENOMEM and GNTST_eagain have the same numeric value. Fix that by turning all grant mapping errors into -ENOENT. This is no problem as all callers of xenbus_map_ring_valloc() only use the return value to print an error message, and in case of mapping errors the grant status value has already been printed by __xenbus_map_ring() before. Signed-off-by: Juergen Gross Reviewed-by: Boris Ostrovsky --- drivers/xen/xenbus/xenbus_client.c | 22 ++++++---------------- 1 file changed, 6 insertions(+), 16 deletions(-) diff --git a/drivers/xen/xenbus/xenbus_client.c b/drivers/xen/xenbus/xenbus_client.c index 9f8372079ecf..4f168b46fbca 100644 --- a/drivers/xen/xenbus/xenbus_client.c +++ b/drivers/xen/xenbus/xenbus_client.c @@ -456,8 +456,7 @@ EXPORT_SYMBOL_GPL(xenbus_free_evtchn); * Map @nr_grefs pages of memory into this domain from another * domain's grant table. xenbus_map_ring_valloc allocates @nr_grefs * pages of virtual address space, maps the pages to that address, and - * sets *vaddr to that address. Returns 0 on success, and GNTST_* - * (see xen/include/interface/grant_table.h) or -ENOMEM / -EINVAL on + * sets *vaddr to that address. Returns 0 on success, and -errno on * error. If an error is returned, device will switch to * XenbusStateClosing and the error message will be saved in XenStore. */ @@ -477,18 +476,11 @@ int xenbus_map_ring_valloc(struct xenbus_device *dev, grant_ref_t *gnt_refs, return -ENOMEM; info->node = kzalloc(sizeof(*info->node), GFP_KERNEL); - if (!info->node) { + if (!info->node) err = -ENOMEM; - goto out; - } - - err = ring_ops->map(dev, info, gnt_refs, nr_grefs, vaddr); - - /* Some hypervisors are buggy and can return 1. */ - if (err > 0) - err = GNTST_general_error; + else + err = ring_ops->map(dev, info, gnt_refs, nr_grefs, vaddr); - out: kfree(info->node); kfree(info); return err; @@ -507,7 +499,6 @@ static int __xenbus_map_ring(struct xenbus_device *dev, bool *leaked) { int i, j; - int err = GNTST_okay; if (nr_grefs > XENBUS_MAX_RING_GRANTS) return -EINVAL; @@ -522,7 +513,6 @@ static int __xenbus_map_ring(struct xenbus_device *dev, for (i = 0; i < nr_grefs; i++) { if (info->map[i].status != GNTST_okay) { - err = info->map[i].status; xenbus_dev_fatal(dev, info->map[i].status, "mapping in shared page %d from domain %d", gnt_refs[i], dev->otherend_id); @@ -531,7 +521,7 @@ static int __xenbus_map_ring(struct xenbus_device *dev, handles[i] = info->map[i].handle; } - return GNTST_okay; + return 0; fail: for (i = j = 0; i < nr_grefs; i++) { @@ -554,7 +544,7 @@ static int __xenbus_map_ring(struct xenbus_device *dev, } } - return err; + return -ENOENT; } /**