| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Thu, 7 Sep 2023 13:03:14 +0300
From: Dan Carpenter <dan.carpenter@...aro.org>
To: oe-kbuild@...ts.linux.dev,
Dmitry Osipenko <dmitry.osipenko@...labora.com>,
David Airlie <airlied@...il.com>,
Gerd Hoffmann <kraxel@...hat.com>,
Gurchetan Singh <gurchetansingh@...omium.org>,
Chia-I Wu <olvaffe@...il.com>, Daniel Vetter <daniel@...ll.ch>,
Maarten Lankhorst <maarten.lankhorst@...ux.intel.com>,
Maxime Ripard <mripard@...nel.org>,
Thomas Zimmermann <tzimmermann@...e.de>,
Christian König <christian.koenig@....com>,
Qiang Yu <yuq825@...il.com>,
Steven Price <steven.price@....com>,
Boris Brezillon <bbrezillon@...nel.org>,
Emma Anholt <emma@...olt.net>, Melissa Wen <mwen@...lia.com>
Cc: lkp@...el.com, oe-kbuild-all@...ts.linux.dev, kernel@...labora.com,
linux-kernel@...r.kernel.org, dri-devel@...ts.freedesktop.org,
virtualization@...ts.linux-foundation.org
Subject: Re: [PATCH v16 15/20] drm/shmem-helper: Add memory shrinker
Hi Dmitry,
kernel test robot noticed the following build warnings:
https://git-scm.com/docs/git-format-patch#_base_tree_information]
url: https://github.com/intel-lab-lkp/linux/commits/Dmitry-Osipenko/drm-shmem-helper-Fix-UAF-in-error-path-when-freeing-SGT-of-imported-GEM/20230904-011037
base: linus/master
patch link: https://lore.kernel.org/r/20230903170736.513347-16-dmitry.osipenko%40collabora.com
patch subject: [PATCH v16 15/20] drm/shmem-helper: Add memory shrinker
config: x86_64-randconfig-161-20230907 (https://download.01.org/0day-ci/archive/20230907/202309070814.jyGJOJzY-lkp@intel.com/config)
compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
reproduce: (https://download.01.org/0day-ci/archive/20230907/202309070814.jyGJOJzY-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@...el.com>
| Reported-by: Dan Carpenter <dan.carpenter@...aro.org>
| Closes: https://lore.kernel.org/r/202309070814.jyGJOJzY-lkp@intel.com/
smatch warnings:
drivers/gpu/drm/drm_gem_shmem_helper.c:733 drm_gem_shmem_fault() error: we previously assumed 'shmem->pages' could be null (see line 724)
vim +733 drivers/gpu/drm/drm_gem_shmem_helper.c
2194a63a818db7 Noralf Trønnes 2019-03-12 702 static vm_fault_t drm_gem_shmem_fault(struct vm_fault *vmf)
2194a63a818db7 Noralf Trønnes 2019-03-12 703 {
2194a63a818db7 Noralf Trønnes 2019-03-12 704 struct vm_area_struct *vma = vmf->vma;
2194a63a818db7 Noralf Trønnes 2019-03-12 705 struct drm_gem_object *obj = vma->vm_private_data;
2194a63a818db7 Noralf Trønnes 2019-03-12 706 struct drm_gem_shmem_object *shmem = to_drm_gem_shmem_obj(obj);
2194a63a818db7 Noralf Trønnes 2019-03-12 707 loff_t num_pages = obj->size >> PAGE_SHIFT;
d611b4a0907cec Neil Roberts 2021-02-23 708 vm_fault_t ret;
2194a63a818db7 Noralf Trønnes 2019-03-12 709 struct page *page;
11d5a4745e00e7 Neil Roberts 2021-02-23 710 pgoff_t page_offset;
2c607edf57db6a Dmitry Osipenko 2023-09-03 711 bool pages_unpinned;
2c607edf57db6a Dmitry Osipenko 2023-09-03 712 int err;
11d5a4745e00e7 Neil Roberts 2021-02-23 713
11d5a4745e00e7 Neil Roberts 2021-02-23 714 /* We don't use vmf->pgoff since that has the fake offset */
11d5a4745e00e7 Neil Roberts 2021-02-23 715 page_offset = (vmf->address - vma->vm_start) >> PAGE_SHIFT;
2194a63a818db7 Noralf Trønnes 2019-03-12 716
21aa27ddc58269 Dmitry Osipenko 2023-05-30 717 dma_resv_lock(shmem->base.resv, NULL);
2194a63a818db7 Noralf Trønnes 2019-03-12 718
2c607edf57db6a Dmitry Osipenko 2023-09-03 719 /* Sanity-check that we have the pages pointer when it should present */
2c607edf57db6a Dmitry Osipenko 2023-09-03 720 pages_unpinned = (shmem->evicted || shmem->madv < 0 ||
2c607edf57db6a Dmitry Osipenko 2023-09-03 721 !refcount_read(&shmem->pages_use_count));
2c607edf57db6a Dmitry Osipenko 2023-09-03 722 drm_WARN_ON_ONCE(obj->dev, !shmem->pages ^ pages_unpinned);
2c607edf57db6a Dmitry Osipenko 2023-09-03 723
2c607edf57db6a Dmitry Osipenko 2023-09-03 @724 if (page_offset >= num_pages || (!shmem->pages && !shmem->evicted)) {
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Should this be || instead of &&? (The other thing that people do is
add "!shmem->evicted" for readability even though it doesn't need to be
checked. So maybe that's the issue, but the checker assumes it needs to
be checked).
d611b4a0907cec Neil Roberts 2021-02-23 725 ret = VM_FAULT_SIGBUS;
d611b4a0907cec Neil Roberts 2021-02-23 726 } else {
2c607edf57db6a Dmitry Osipenko 2023-09-03 727 err = drm_gem_shmem_swapin_locked(shmem);
Or maybe it's because the kbuild bot can't use the cross function db
and shmem->pages is assigned here?
2c607edf57db6a Dmitry Osipenko 2023-09-03 728 if (err) {
2c607edf57db6a Dmitry Osipenko 2023-09-03 729 ret = VM_FAULT_OOM;
2c607edf57db6a Dmitry Osipenko 2023-09-03 730 goto unlock;
2c607edf57db6a Dmitry Osipenko 2023-09-03 731 }
2c607edf57db6a Dmitry Osipenko 2023-09-03 732
11d5a4745e00e7 Neil Roberts 2021-02-23 @733 page = shmem->pages[page_offset];
^^^^^^^^^^^^
Unchecked dereference
2194a63a818db7 Noralf Trønnes 2019-03-12 734
8b93d1d7dbd578 Daniel Vetter 2021-08-12 735 ret = vmf_insert_pfn(vma, vmf->address, page_to_pfn(page));
d611b4a0907cec Neil Roberts 2021-02-23 736 }
d611b4a0907cec Neil Roberts 2021-02-23 737
2c607edf57db6a Dmitry Osipenko 2023-09-03 738 unlock:
21aa27ddc58269 Dmitry Osipenko 2023-05-30 739 dma_resv_unlock(shmem->base.resv);
d611b4a0907cec Neil Roberts 2021-02-23 740
d611b4a0907cec Neil Roberts 2021-02-23 741 return ret;
2194a63a818db7 Noralf Trønnes 2019-03-12 742 }
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
Powered by blists - more mailing lists