mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Dawei Li <set_pte_at@outlook.com>
To: zackr@vmware.com, airlied@gmail.com, daniel@ffwll.ch
Cc: krastevm@vmware.com, linux-graphics-maintainer@vmware.com,
	dri-devel@lists.freedesktop.org, linux-kernel@vger.kernel.org,
	Dawei Li <set_pte_at@outlook.com>
Subject: [PATCH v2] drm/vmwgfx: Protect pin_user_pages with mmap_lock
Date: Sun,  6 Nov 2022 23:47:36 +0800	[thread overview]
Message-ID: <TYCP286MB23235B138D18ECA0797A6D0FCA3D9@TYCP286MB2323.JPNP286.PROD.OUTLOOK.COM> (raw)

This patch includes changes below:
1) pin_user_pages() is unsafe without protection of mmap_lock,
   fix it by calling mmap_read_lock() & mmap_read_unlock().
2) fix & refactor the incorrect exception handling procedure in
   vmw_mksstat_add_ioctl().

Fixes: 7a7a933edd6c ("drm/vmwgfx: Introduce VMware mks-guest-stats")
Signed-off-by: Dawei Li <set_pte_at@outlook.com>
---
v1:
https://lore.kernel.org/all/TYCP286MB23235C9A9FCF85C045F95EA7CA4F9@TYCP286MB2323.JPNP286.PROD.OUTLOOK.COM/

v1->v2:
Rebased to latest vmwgfx/drm-misc-fixes
---
 drivers/gpu/drm/vmwgfx/vmwgfx_msg.c | 23 ++++++++++++++---------
 1 file changed, 14 insertions(+), 9 deletions(-)

diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
index 089046fa21be..ec40a3364e0a 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_msg.c
@@ -1020,9 +1020,9 @@ int vmw_mksstat_add_ioctl(struct drm_device *dev, void *data,
 	const size_t num_pages_info = PFN_UP(arg->info_len);
 	const size_t num_pages_strs = PFN_UP(arg->strs_len);
 	long desc_len;
-	long nr_pinned_stat;
-	long nr_pinned_info;
-	long nr_pinned_strs;
+	long nr_pinned_stat = 0;
+	long nr_pinned_info = 0;
+	long nr_pinned_strs = 0;
 	struct page *pages_stat[ARRAY_SIZE(pdesc->statPPNs)];
 	struct page *pages_info[ARRAY_SIZE(pdesc->infoPPNs)];
 	struct page *pages_strs[ARRAY_SIZE(pdesc->strsPPNs)];
@@ -1084,28 +1084,33 @@ int vmw_mksstat_add_ioctl(struct drm_device *dev, void *data,
 	reset_ppn_array(pdesc->infoPPNs, ARRAY_SIZE(pdesc->infoPPNs));
 	reset_ppn_array(pdesc->strsPPNs, ARRAY_SIZE(pdesc->strsPPNs));
 
+	/* pin_user_pages() needs protection of mmap_lock */
+	mmap_read_lock(current->mm);
+
 	/* Pin mksGuestStat user pages and store those in the instance descriptor */
 	nr_pinned_stat = pin_user_pages(arg->stat, num_pages_stat, FOLL_LONGTERM, pages_stat, NULL);
 	if (num_pages_stat != nr_pinned_stat)
-		goto err_pin_stat;
+		goto __err_pin_pages;
 
 	for (i = 0; i < num_pages_stat; ++i)
 		pdesc->statPPNs[i] = page_to_pfn(pages_stat[i]);
 
 	nr_pinned_info = pin_user_pages(arg->info, num_pages_info, FOLL_LONGTERM, pages_info, NULL);
 	if (num_pages_info != nr_pinned_info)
-		goto err_pin_info;
+		goto __err_pin_pages;
 
 	for (i = 0; i < num_pages_info; ++i)
 		pdesc->infoPPNs[i] = page_to_pfn(pages_info[i]);
 
 	nr_pinned_strs = pin_user_pages(arg->strs, num_pages_strs, FOLL_LONGTERM, pages_strs, NULL);
 	if (num_pages_strs != nr_pinned_strs)
-		goto err_pin_strs;
+		goto __err_pin_pages;
 
 	for (i = 0; i < num_pages_strs; ++i)
 		pdesc->strsPPNs[i] = page_to_pfn(pages_strs[i]);
 
+	mmap_read_unlock(current->mm);
+
 	/* Send the descriptor to the host via a hypervisor call. The mksGuestStat
 	   pages will remain in use until the user requests a matching remove stats
 	   or a stats reset occurs. */
@@ -1120,15 +1125,15 @@ int vmw_mksstat_add_ioctl(struct drm_device *dev, void *data,
 
 	return 0;
 
-err_pin_strs:
+__err_pin_pages:
+	mmap_read_unlock(current->mm);
+
 	if (nr_pinned_strs > 0)
 		unpin_user_pages(pages_strs, nr_pinned_strs);
 
-err_pin_info:
 	if (nr_pinned_info > 0)
 		unpin_user_pages(pages_info, nr_pinned_info);
 
-err_pin_stat:
 	if (nr_pinned_stat > 0)
 		unpin_user_pages(pages_stat, nr_pinned_stat);
 
-- 
2.25.1


             reply	other threads:[~2022-11-06 15:48 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-11-06 15:47 Dawei Li [this message]
2022-11-07 15:16 ` Martin Krastev (VMware)

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=TYCP286MB23235B138D18ECA0797A6D0FCA3D9@TYCP286MB2323.JPNP286.PROD.OUTLOOK.COM \
    --to=set_pte_at@outlook.com \
    --cc=airlied@gmail.com \
    --cc=daniel@ffwll.ch \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=krastevm@vmware.com \
    --cc=linux-graphics-maintainer@vmware.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=zackr@vmware.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®