From: Zheng Wang zyytlz.wz@163.com
mainline inclusion from mainline-v6.2-rc3 commit 4a61648af68f5ba4884f0e3b494ee1cabc4b6620 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I5XXFF CVE: CVE-2022-3707
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...
--------------------------------
If intel_gvt_dma_map_guest_page failed, it will call ppgtt_invalidate_spt, which will finally free the spt. But the caller function ppgtt_populate_spt_by_guest_entry does not notice that, it will free spt again in its error path.
Fix this by canceling the mapping of DMA address and freeing sub_spt. Besides, leave the handle of spt destroy to caller function instead of callee function when error occurs.
Fixes: b901b252b6cf ("drm/i915/gvt: Add 2M huge gtt support") Signed-off-by: Zheng Wang zyytlz.wz@163.com Reviewed-by: Zhenyu Wang zhenyuw@linux.intel.com Signed-off-by: Zhenyu Wang zhenyuw@linux.intel.com Link: http://patchwork.freedesktop.org/patch/msgid/20221229165641.1192455-1-zyytlz... conflicts: drivers/gpu/drm/i915/gvt/gtt.c Signed-off-by: Wang ShaoBo bobo.shaobowang@huawei.com Reviewed-by: Xiu Jianfeng xiujianfeng@huawei.com Reviewed-by: Wei Li liwei391@huawei.com Signed-off-by: Yongqiang Liu liuyongqiang13@huawei.com --- drivers/gpu/drm/i915/gvt/gtt.c | 17 +++++++++++++---- 1 file changed, 13 insertions(+), 4 deletions(-)
diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c index 40b32b4d1d98..afbc648befec 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.c +++ b/drivers/gpu/drm/i915/gvt/gtt.c @@ -1155,10 +1155,8 @@ static int split_2MB_gtt_entry(struct intel_vgpu *vgpu, for_each_shadow_entry(sub_spt, &sub_se, sub_index) { ret = intel_gvt_hypervisor_dma_map_guest_page(vgpu, start_gfn + sub_index, PAGE_SIZE, &dma_addr); - if (ret) { - ppgtt_invalidate_spt(spt); - return ret; - } + if (ret) + goto err; sub_se.val64 = se->val64;
/* Copy the PAT field from PDE. */ @@ -1177,6 +1175,17 @@ static int split_2MB_gtt_entry(struct intel_vgpu *vgpu, ops->set_pfn(se, sub_spt->shadow_page.mfn); ppgtt_set_shadow_entry(spt, se, index); return 0; +err: + /* Cancel the existing addess mappings of DMA addr. */ + for_each_present_shadow_entry(sub_spt, &sub_se, sub_index) { + gvt_vdbg_mm("invalidate 4K entry\n"); + ppgtt_invalidate_pte(sub_spt, &sub_se); + } + /* Release the new allocated spt. */ + trace_spt_change(sub_spt->vgpu->id, "release", sub_spt, + sub_spt->guest_page.gfn, sub_spt->shadow_page.type); + ppgtt_free_spt(sub_spt); + return ret; }
static int split_64KB_gtt_entry(struct intel_vgpu *vgpu,
From: Thiago Rafael Becker trbecker@gmail.com
mainline inclusion from mainline-v5.16-rc6 commit a31080899d5fdafcccf7f39dd214a814a2c82626 category: bugfix bugzilla: https://gitee.com/openeuler/kernel/issues/I6D6RL CVE: NA
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i...
--------------------------------
mount.cifs can pass a device with multiple delimiters in it. This will cause rename(2) to fail with ENOENT.
V2: - Make sanitize_path more readable. - Fix multiple delimiters between UNC and prepath. - Avoid a memory leak if a bad user starts putting a lot of delimiters in the path on purpose.
BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=2031200 Fixes: 24e0a1eff9e2 ("cifs: switch to new mount api") Cc: stable@vger.kernel.org # 5.11+ Acked-by: Ronnie Sahlberg lsahlber@redhat.com Signed-off-by: Thiago Rafael Becker trbecker@gmail.com Signed-off-by: Steve French stfrench@microsoft.com
conflicts: fs/cifs/fs_context.c
Signed-off-by: ZhaoLong Wang wangzhaolong1@huawei.com Reviewed-by: Zhang Xiaoxu zhangxiaoxu5@huawei.com Signed-off-by: Yongqiang Liu liuyongqiang13@huawei.com --- fs/cifs/connect.c | 38 +++++++++++++++++++++++++++++++++++++- 1 file changed, 37 insertions(+), 1 deletion(-)
diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 976aae1eaf73..4ca6fd66df03 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -1372,6 +1372,42 @@ cifs_parse_smb_version(char *value, struct smb_vol *vol, bool is_smb3) return 0; }
+/* + * Remove duplicate path delimiters. Windows is supposed to do that + * but there are some bugs that prevent rename from working if there are + * multiple delimiters. + * + * Returns a sanitized duplicate of @path. The caller is responsible for + * cleaning up the original. + */ +#define IS_DELIM(c) ((c) == '/' || (c) == '\') +static char *sanitize_path(char *path) +{ + char *cursor1 = path, *cursor2 = path; + + /* skip all prepended delimiters */ + while (IS_DELIM(*cursor1)) + cursor1++; + + /* copy the first letter */ + *cursor2 = *cursor1; + + /* copy the remainder... */ + while (*(cursor1++)) { + /* ... skipping all duplicated delimiters */ + if (IS_DELIM(*cursor1) && IS_DELIM(*cursor2)) + continue; + *(++cursor2) = *cursor1; + } + + /* if the last character is a delimiter, skip it */ + if (IS_DELIM(*(cursor2 - 1))) + cursor2--; + + *(cursor2) = '\0'; + return kstrdup(path, GFP_KERNEL); +} + /* * Parse a devname into substrings and populate the vol->UNC and vol->prepath * fields with the result. Returns 0 on success and an error otherwise. @@ -1420,7 +1456,7 @@ cifs_parse_devname(const char *devname, struct smb_vol *vol) if (!*pos) return 0;
- vol->prepath = kstrdup(pos, GFP_KERNEL); + vol->prepath = sanitize_path(pos); if (!vol->prepath) return -ENOMEM;
-
Yongqiang Liu