The patch titled
Subject: mm: kmsan: handle alloc failures in kmsan_vmap_pages_range_noflush()
has been added to the -mm mm-hotfixes-unstable branch. Its filename is
mm-kmsan-handle-alloc-failures-in-kmsan_vmap_pages_range_noflush.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patche…
This patch will later appear in the mm-hotfixes-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: Alexander Potapenko <glider(a)google.com>
Subject: mm: kmsan: handle alloc failures in kmsan_vmap_pages_range_noflush()
Date: Wed, 12 Apr 2023 16:52:59 +0200
As reported by Dipanjan Das, when KMSAN is used together with kernel fault
injection (or, generally, even without the latter), calls to kcalloc() or
__vmap_pages_range_noflush() may fail, leaving the metadata mappings for
the virtual mapping in an inconsistent state. When these metadata
mappings are accessed later, the kernel crashes.
To address the problem, we return a non-zero error code from
kmsan_vmap_pages_range_noflush() in the case of any allocation/mapping
failure inside it, and make vmap_pages_range_noflush() return an error if
KMSAN fails to allocate the metadata.
This patch also removes KMSAN_WARN_ON() from vmap_pages_range_noflush(),
as these allocation failures are not fatal anymore.
Link: https://lkml.kernel.org/r/20230412145300.3651840-1-glider@google.com
Fixes: b073d7f8aee4 ("mm: kmsan: maintain KMSAN metadata for page operations")
Signed-off-by: Alexander Potapenko <glider(a)google.com>
Reported-by: Dipanjan Das <mail.dipanjan.das(a)gmail.com>
Link: https://lore.kernel.org/linux-mm/CANX2M5ZRrRA64k0hOif02TjmY9kbbO2aCBPyq79es…
Cc: <stable(a)vger.kernel.org>
Cc: Christoph Hellwig <hch(a)infradead.org>
Cc: Dmitry Vyukov <dvyukov(a)google.com>
Cc: Marco Elver <elver(a)google.com>
Cc: Uladzislau Rezki (Sony) <urezki(a)gmail.com>
Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org>
---
include/linux/kmsan.h | 19 ++++++++++---------
mm/kmsan/shadow.c | 27 ++++++++++++++++++---------
mm/vmalloc.c | 6 +++++-
3 files changed, 33 insertions(+), 19 deletions(-)
--- a/include/linux/kmsan.h~mm-kmsan-handle-alloc-failures-in-kmsan_vmap_pages_range_noflush
+++ a/include/linux/kmsan.h
@@ -134,11 +134,12 @@ void kmsan_kfree_large(const void *ptr);
* @page_shift: page_shift passed to vmap_range_noflush().
*
* KMSAN maps shadow and origin pages of @pages into contiguous ranges in
- * vmalloc metadata address range.
+ * vmalloc metadata address range. Returns 0 on success, callers must check
+ * for non-zero return value.
*/
-void kmsan_vmap_pages_range_noflush(unsigned long start, unsigned long end,
- pgprot_t prot, struct page **pages,
- unsigned int page_shift);
+int kmsan_vmap_pages_range_noflush(unsigned long start, unsigned long end,
+ pgprot_t prot, struct page **pages,
+ unsigned int page_shift);
/**
* kmsan_vunmap_kernel_range_noflush() - Notify KMSAN about a vunmap.
@@ -281,11 +282,11 @@ static inline void kmsan_kfree_large(con
{
}
-static inline void kmsan_vmap_pages_range_noflush(unsigned long start,
- unsigned long end,
- pgprot_t prot,
- struct page **pages,
- unsigned int page_shift)
+static inline int kmsan_vmap_pages_range_noflush(unsigned long start,
+ unsigned long end,
+ pgprot_t prot,
+ struct page **pages,
+ unsigned int page_shift)
{
}
--- a/mm/kmsan/shadow.c~mm-kmsan-handle-alloc-failures-in-kmsan_vmap_pages_range_noflush
+++ a/mm/kmsan/shadow.c
@@ -216,27 +216,29 @@ void kmsan_free_page(struct page *page,
kmsan_leave_runtime();
}
-void kmsan_vmap_pages_range_noflush(unsigned long start, unsigned long end,
- pgprot_t prot, struct page **pages,
- unsigned int page_shift)
+int kmsan_vmap_pages_range_noflush(unsigned long start, unsigned long end,
+ pgprot_t prot, struct page **pages,
+ unsigned int page_shift)
{
unsigned long shadow_start, origin_start, shadow_end, origin_end;
struct page **s_pages, **o_pages;
- int nr, mapped;
+ int nr, mapped, err = 0;
if (!kmsan_enabled)
- return;
+ return 0;
shadow_start = vmalloc_meta((void *)start, KMSAN_META_SHADOW);
shadow_end = vmalloc_meta((void *)end, KMSAN_META_SHADOW);
if (!shadow_start)
- return;
+ return 0;
nr = (end - start) / PAGE_SIZE;
s_pages = kcalloc(nr, sizeof(*s_pages), GFP_KERNEL);
o_pages = kcalloc(nr, sizeof(*o_pages), GFP_KERNEL);
- if (!s_pages || !o_pages)
+ if (!s_pages || !o_pages) {
+ err = -ENOMEM;
goto ret;
+ }
for (int i = 0; i < nr; i++) {
s_pages[i] = shadow_page_for(pages[i]);
o_pages[i] = origin_page_for(pages[i]);
@@ -249,10 +251,16 @@ void kmsan_vmap_pages_range_noflush(unsi
kmsan_enter_runtime();
mapped = __vmap_pages_range_noflush(shadow_start, shadow_end, prot,
s_pages, page_shift);
- KMSAN_WARN_ON(mapped);
+ if (mapped) {
+ err = mapped;
+ goto ret;
+ }
mapped = __vmap_pages_range_noflush(origin_start, origin_end, prot,
o_pages, page_shift);
- KMSAN_WARN_ON(mapped);
+ if (mapped) {
+ err = mapped;
+ goto ret;
+ }
kmsan_leave_runtime();
flush_tlb_kernel_range(shadow_start, shadow_end);
flush_tlb_kernel_range(origin_start, origin_end);
@@ -262,6 +270,7 @@ void kmsan_vmap_pages_range_noflush(unsi
ret:
kfree(s_pages);
kfree(o_pages);
+ return err;
}
/* Allocate metadata for pages allocated at boot time. */
--- a/mm/vmalloc.c~mm-kmsan-handle-alloc-failures-in-kmsan_vmap_pages_range_noflush
+++ a/mm/vmalloc.c
@@ -605,7 +605,11 @@ int __vmap_pages_range_noflush(unsigned
int vmap_pages_range_noflush(unsigned long addr, unsigned long end,
pgprot_t prot, struct page **pages, unsigned int page_shift)
{
- kmsan_vmap_pages_range_noflush(addr, end, prot, pages, page_shift);
+ int ret = kmsan_vmap_pages_range_noflush(addr, end, prot, pages,
+ page_shift);
+
+ if (ret)
+ return ret;
return __vmap_pages_range_noflush(addr, end, prot, pages, page_shift);
}
_
Patches currently in -mm which might be from glider(a)google.com are
mm-kmsan-handle-alloc-failures-in-kmsan_vmap_pages_range_noflush.patch
mm-kmsan-handle-alloc-failures-in-kmsan_ioremap_page_range.patch
The patch titled
Subject: Revert "userfaultfd: don't fail on unrecognized features"
has been added to the -mm mm-hotfixes-unstable branch. Its filename is
revert-userfaultfd-dont-fail-on-unrecognized-features.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patche…
This patch will later appear in the mm-hotfixes-unstable branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via the mm-everything
branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there every 2-3 working days
------------------------------------------------------
From: Peter Xu <peterx(a)redhat.com>
Subject: Revert "userfaultfd: don't fail on unrecognized features"
Date: Wed, 12 Apr 2023 12:38:52 -0400
This is a proposal to revert commit 914eedcb9ba0ff53c33808.
I found this when writing a simple UFFDIO_API test to be the first unit
test in this set. Two things breaks with the commit:
- UFFDIO_API check was lost and missing. According to man page, the
kernel should reject ioctl(UFFDIO_API) if uffdio_api.api != 0xaa. This
check is needed if the api version will be extended in the future, or
user app won't be able to identify which is a new kernel.
- Feature flags checks were removed, which means UFFDIO_API with a
feature that does not exist will also succeed. According to the man
page, we should (and it makes sense) to reject ioctl(UFFDIO_API) if
unknown features passed in.
Link: https://lore.kernel.org/r/20220722201513.1624158-1-axelrasmussen@google.com
Link: https://lkml.kernel.org/r/20230412163922.327282-2-peterx@redhat.com
Fixes: 914eedcb9ba0 ("userfaultfd: don't fail on unrecognized features")
Signed-off-by: Peter Xu <peterx(a)redhat.com>
Acked-by: David Hildenbrand <david(a)redhat.com>
Cc: Axel Rasmussen <axelrasmussen(a)google.com>
Cc: Dmitry Safonov <0x7f454c46(a)gmail.com>
Cc: Mike Kravetz <mike.kravetz(a)oracle.com>
Cc: Mike Rapoport (IBM) <rppt(a)kernel.org>
Cc: Zach O'Keefe <zokeefe(a)google.com>
Cc: <stable(a)vger.kernel.org>
Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org>
---
fs/userfaultfd.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
--- a/fs/userfaultfd.c~revert-userfaultfd-dont-fail-on-unrecognized-features
+++ a/fs/userfaultfd.c
@@ -1955,8 +1955,10 @@ static int userfaultfd_api(struct userfa
ret = -EFAULT;
if (copy_from_user(&uffdio_api, buf, sizeof(uffdio_api)))
goto out;
- /* Ignore unsupported features (userspace built against newer kernel) */
- features = uffdio_api.features & UFFD_API_FEATURES;
+ features = uffdio_api.features;
+ ret = -EINVAL;
+ if (uffdio_api.api != UFFD_API || (features & ~UFFD_API_FEATURES))
+ goto err_out;
ret = -EPERM;
if ((features & UFFD_FEATURE_EVENT_FORK) && !capable(CAP_SYS_PTRACE))
goto err_out;
_
Patches currently in -mm which might be from peterx(a)redhat.com are
mm-khugepaged-check-again-on-anon-uffd-wp-during-isolation.patch
revert-userfaultfd-dont-fail-on-unrecognized-features.patch
selftests-mm-update-gitignore-with-two-missing-tests.patch
selftests-mm-dump-a-summary-in-run_vmtestssh.patch
selftests-mm-merge-utilh-into-vm_utilh.patch
selftests-mm-use-test_gen_progs-where-proper.patch
selftests-mm-link-vm_utilc-always.patch
selftests-mm-merge-default_huge_page_size-into-one.patch
selftests-mm-use-pm_-macros-in-vm_utilsh.patch
selftests-mm-reuse-pagemap_get_entry-in-vm_utilh.patch
selftests-mm-test-uffdio_zeropage-only-when-hugetlb.patch
selftests-mm-drop-test_uffdio_zeropage_eexist.patch
selftests-mm-create-uffd-common.patch
selftests-mm-split-uffd-tests-into-uffd-stress-and-uffd-unit-tests.patch
selftests-mm-uffd_register.patch
selftests-mm-uffd_open_devsys.patch
selftests-mm-uffdio_api-test.patch
selftests-mm-drop-global-mem_fd-in-uffd-tests.patch
selftests-mm-drop-global-hpage_size-in-uffd-tests.patch
selftests-mm-rename-uffd_stats-to-uffd_args.patch
selftests-mm-let-uffd_handle_page_fault-take-wp-parameter.patch
selftests-mm-allow-allocate_area-to-fail-properly.patch
selftests-mm-add-framework-for-uffd-unit-test.patch
selftests-mm-move-uffd-pagemap-test-to-unit-test.patch
selftests-mm-move-uffd-minor-test-to-unit-test.patch
selftests-mm-move-uffd-sig-events-tests-into-uffd-unit-tests.patch
selftests-mm-move-zeropage-test-into-uffd-unit-tests.patch
selftests-mm-workaround-no-way-to-detect-uffd-minor-wp.patch
selftests-mm-allow-uffd-test-to-skip-properly-with-no-privilege.patch
selftests-mm-drop-sys-dev-test-in-uffd-stress-test.patch
selftests-mm-add-shmem-private-test-to-uffd-stress.patch
selftests-mm-add-uffdio-register-ioctls-test.patch
This is a regressoin introduced in b07db3958485 ("fbcon: Ditch error
handling for con2fb_release_oldinfo"). I failed to realize that the
if (!err) checks and therefore flattening the control flow due to
con2fb_release_oldinfo() (which was impossible) wasn't correct,
because con2fb_acquire_newinfo() could fail.
Fix this with an early return statement.
Note that there's still a difference compared to the orginal state of
the code, the below lines are now also skipped on error:
if (!search_fb_in_map(info_idx))
info_idx = newidx;
These are only needed when we've actually thrown out an old fb_info
from the console mappings, which only happens later on.
Also move the fbcon_add_cursor_work() call into the same if block,
it's all protected by console_lock so doesn't matter when we set up
the blinking cursor delayed work anyway. This further simplifies the
control flow and allows us to ditch the found local variable.
Signed-off-by: Daniel Vetter <daniel.vetter(a)intel.com>
Fixes: b07db3958485 ("fbcon: Ditch error handling for con2fb_release_oldinfo")
Cc: Thomas Zimmermann <tzimmermann(a)suse.de>
Cc: Sam Ravnborg <sam(a)ravnborg.org>
Cc: Xingyuan Mo <hdthky0(a)gmail.com>
Cc: Thomas Zimmermann <tzimmermann(a)suse.de>
Cc: Helge Deller <deller(a)gmx.de>
Cc: <stable(a)vger.kernel.org> # v5.19+
---
drivers/video/fbdev/core/fbcon.c | 17 ++++++++---------
1 file changed, 8 insertions(+), 9 deletions(-)
diff --git a/drivers/video/fbdev/core/fbcon.c b/drivers/video/fbdev/core/fbcon.c
index 0a2c47df01f4..52f6ce8e794a 100644
--- a/drivers/video/fbdev/core/fbcon.c
+++ b/drivers/video/fbdev/core/fbcon.c
@@ -823,7 +823,7 @@ static int set_con2fb_map(int unit, int newidx, int user)
int oldidx = con2fb_map[unit];
struct fb_info *info = fbcon_registered_fb[newidx];
struct fb_info *oldinfo = NULL;
- int found, err = 0, show_logo;
+ int err = 0, show_logo;
WARN_CONSOLE_UNLOCKED();
@@ -841,26 +841,25 @@ static int set_con2fb_map(int unit, int newidx, int user)
if (oldidx != -1)
oldinfo = fbcon_registered_fb[oldidx];
- found = search_fb_in_map(newidx);
-
- if (!err && !found) {
+ if (!search_fb_in_map(newidx)) {
err = con2fb_acquire_newinfo(vc, info, unit);
- if (!err)
- con2fb_map[unit] = newidx;
+ if (err)
+ return err;
+
+ con2fb_map[unit] = newidx;
+ fbcon_add_cursor_work(info);
}
/*
* If old fb is not mapped to any of the consoles,
* fbcon should release it.
*/
- if (!err && oldinfo && !search_fb_in_map(oldidx))
+ if (oldinfo && !search_fb_in_map(oldidx))
con2fb_release_oldinfo(vc, oldinfo, info);
show_logo = (fg_console == 0 && !user &&
logo_shown != FBCON_LOGO_DONTSHOW);
- if (!found)
- fbcon_add_cursor_work(info);
con2fb_map_boot[unit] = newidx;
con2fb_init_display(vc, info, unit, show_logo);
--
2.40.0
This is a proposal to revert commit 914eedcb9ba0ff53c33808.
I found this when writting a simple UFFDIO_API test to be the first unit
test in this set. Two things breaks with the commit:
- UFFDIO_API check was lost and missing. According to man page, the
kernel should reject ioctl(UFFDIO_API) if uffdio_api.api != 0xaa. This
check is needed if the api version will be extended in the future, or
user app won't be able to identify which is a new kernel.
- Feature flags checks were removed, which means UFFDIO_API with a
feature that does not exist will also succeed. According to the man
page, we should (and it makes sense) to reject ioctl(UFFDIO_API) if
unknown features passed in.
Link: https://lore.kernel.org/r/20220722201513.1624158-1-axelrasmussen@google.com
Fixes: 914eedcb9ba0 ("userfaultfd: don't fail on unrecognized features")
Cc: Axel Rasmussen <axelrasmussen(a)google.com>
Cc: Dmitry Safonov <0x7f454c46(a)gmail.com>
Cc: Mike Rapoport <rppt(a)linux.vnet.ibm.com>
Cc: linux-stable <stable(a)vger.kernel.org>
Acked-by: David Hildenbrand <david(a)redhat.com>
Signed-off-by: Peter Xu <peterx(a)redhat.com>
---
fs/userfaultfd.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c
index 8395605790f6..3b2a41c330e6 100644
--- a/fs/userfaultfd.c
+++ b/fs/userfaultfd.c
@@ -1977,8 +1977,10 @@ static int userfaultfd_api(struct userfaultfd_ctx *ctx,
ret = -EFAULT;
if (copy_from_user(&uffdio_api, buf, sizeof(uffdio_api)))
goto out;
- /* Ignore unsupported features (userspace built against newer kernel) */
- features = uffdio_api.features & UFFD_API_FEATURES;
+ features = uffdio_api.features;
+ ret = -EINVAL;
+ if (uffdio_api.api != UFFD_API || (features & ~UFFD_API_FEATURES))
+ goto err_out;
ret = -EPERM;
if ((features & UFFD_FEATURE_EVENT_FORK) && !capable(CAP_SYS_PTRACE))
goto err_out;
--
2.39.1