From: "Pedro Falcato" <pedro.falcato@gmail.com>
To: devel@edk2.groups.io
Cc: Pedro Falcato <pedro.falcato@gmail.com>,
Leif Lindholm <leif@nuviainc.com>,
Michael D Kinney <michael.d.kinney@intel.com>,
Bret Barkelew <Bret.Barkelew@microsoft.com>
Subject: [edk2-platforms PATCH v2 5/5] Ext4Pkg: Sanity check more EXT4_DIR_ENTRY values.
Date: Sat, 21 Aug 2021 15:47:10 +0100 [thread overview]
Message-ID: <20210821144711.39546-6-pedro.falcato@gmail.com> (raw)
In-Reply-To: <20210821144711.39546-1-pedro.falcato@gmail.com>
This should close up some possible exploits using crafted
filesystem images.
Cc: Leif Lindholm <leif@nuviainc.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Cc: Bret Barkelew <Bret.Barkelew@microsoft.com>
Signed-off-by: Pedro Falcato <pedro.falcato@gmail.com>
---
Features/Ext4Pkg/Ext4Dxe/Directory.c | 90 ++++++++++++++++------------
1 file changed, 51 insertions(+), 39 deletions(-)
diff --git a/Features/Ext4Pkg/Ext4Dxe/Directory.c b/Features/Ext4Pkg/Ext4Dxe/Directory.c
index 7d1b2dcfe524..102c82f05da0 100644
--- a/Features/Ext4Pkg/Ext4Dxe/Directory.c
+++ b/Features/Ext4Pkg/Ext4Dxe/Directory.c
@@ -50,6 +50,37 @@ Ext4GetUcs2DirentName (
return Status;
}
+/**
+ Validates a directory entry.
+
+ @param[in] Dirent Pointer to the directory entry.
+
+ @retval TRUE Valid directory entry.
+ FALSE Invalid directory entry.
+**/
+STATIC
+BOOLEAN
+Ext4ValidDirent (
+ IN CONST EXT4_DIR_ENTRY *Dirent
+ )
+{
+ UINTN RequiredSize;
+
+ RequiredSize = Dirent->name_len + EXT4_MIN_DIR_ENTRY_LEN;
+
+ if (Dirent->rec_len < RequiredSize) {
+ DEBUG ((DEBUG_ERROR, "[ext4] dirent size %lu too small (compared to %lu)\n", Dirent->rec_len, RequiredSize));
+ return FALSE;
+ }
+
+ // Dirent sizes need to be 4 byte aligned
+ if (Dirent->rec_len % 4) {
+ return FALSE;
+ }
+
+ return TRUE;
+}
+
/**
Retrieves a directory entry.
@@ -75,11 +106,11 @@ Ext4RetrieveDirent (
UINT64 DirInoSize;
UINT32 BlockRemainder;
UINTN Length;
- CHAR8 *BufPtr;
EXT4_DIR_ENTRY *Entry;
UINTN RemainingBlock;
CHAR16 DirentUcs2Name[EXT4_NAME_MAX + 1];
UINTN ToCopy;
+ UINTN BlockOffset;
Status = EFI_NOT_FOUND;
Buf = AllocatePool (Partition->BlockSize);
@@ -109,14 +140,19 @@ Ext4RetrieveDirent (
return Status;
}
- for (BufPtr = Buf; BufPtr < Buf + Partition->BlockSize; ) {
- Entry = (EXT4_DIR_ENTRY *)BufPtr;
- if (Entry->rec_len == 0) {
+ for (BlockOffset = 0; BlockOffset < Partition->BlockSize; ) {
+ Entry = (EXT4_DIR_ENTRY *)(Buf + BlockOffset);
+ RemainingBlock = Partition->BlockSize - BlockOffset;
+ // Check if the minimum directory entry fits inside [BlockOffset, EndOfBlock]
+ if (RemainingBlock < EXT4_MIN_DIR_ENTRY_LEN) {
FreePool (Buf);
return EFI_VOLUME_CORRUPTED;
}
- RemainingBlock = Partition->BlockSize - (BufPtr - Buf);
+ if (!Ext4ValidDirent (Entry)) {
+ FreePool (Buf);
+ return EFI_VOLUME_CORRUPTED;
+ }
if (Entry->name_len > RemainingBlock || Entry->rec_len > RemainingBlock) {
// Corrupted filesystem
@@ -131,12 +167,13 @@ Ext4RetrieveDirent (
2) Linux and a number of BSDs also have a filename limit of 255.
*/
if (Entry->name_len > EXT4_NAME_MAX) {
+ BlockOffset += Entry->rec_len;
continue;
}
// Unused entry
if (Entry->inode == 0) {
- BufPtr += Entry->rec_len;
+ BlockOffset += Entry->rec_len;
continue;
}
@@ -150,7 +187,7 @@ Ext4RetrieveDirent (
if (EFI_ERROR (Status)) {
// If we error out, skip this entry
// I'm not sure if this is correct behaviour, but I don't think there's a precedent here.
- BufPtr += Entry->rec_len;
+ BlockOffset += Entry->rec_len;
continue;
}
@@ -163,7 +200,7 @@ Ext4RetrieveDirent (
return EFI_SUCCESS;
}
- BufPtr += Entry->rec_len;
+ BlockOffset += Entry->rec_len;
}
Off += Partition->BlockSize;
@@ -379,37 +416,6 @@ Ext4OpenVolume (
return EFI_SUCCESS;
}
-/**
- Validates a directory entry.
-
- @param[in] Dirent Pointer to the directory entry.
-
- @retval TRUE Valid directory entry.
- FALSE Invalid directory entry.
-**/
-STATIC
-BOOLEAN
-Ext4ValidDirent (
- IN CONST EXT4_DIR_ENTRY *Dirent
- )
-{
- UINTN RequiredSize;
-
- RequiredSize = Dirent->name_len + EXT4_MIN_DIR_ENTRY_LEN;
-
- if (Dirent->rec_len < RequiredSize) {
- DEBUG ((DEBUG_ERROR, "[ext4] dirent size %lu too small (compared to %lu)\n", Dirent->rec_len, RequiredSize));
- return FALSE;
- }
-
- // Dirent sizes need to be 4 byte aligned
- if (Dirent->rec_len % 4) {
- return FALSE;
- }
-
- return TRUE;
-}
-
/**
Reads a directory entry.
@@ -481,6 +487,12 @@ Ext4ReadDir (
goto Out;
}
+ // Check if the entire dir entry length fits in Len
+ if (Len < EXT4_MIN_DIR_ENTRY_LEN + Entry.name_len) {
+ Status = EFI_VOLUME_CORRUPTED;
+ goto Out;
+ }
+
// We don't care about passing . or .. entries to the caller of ReadDir(),
// since they're generally useless entries *and* may break things if too
// many callers assume FAT32.
--
2.33.0
next prev parent reply other threads:[~2021-08-21 14:48 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-21 14:47 [edk2-platforms PATCH v2 0/5] Ext4Pkg: Fix bugs Pedro Falcato
2021-08-21 14:47 ` [edk2-platforms PATCH v2 1/5] Ext4Pkg: Fix incorrect usage of Ext4InitExtentsMap Pedro Falcato
2021-08-21 14:47 ` [edk2-platforms PATCH v2 2/5] Ext4Pkg: Hide "." and ".." entries from Read() callers Pedro Falcato
2021-08-21 14:47 ` [edk2-platforms PATCH v2 3/5] Ext4Pkg: Add a directory entry tree Pedro Falcato
2021-08-21 14:47 ` [edk2-platforms PATCH v2 4/5] Ext4Pkg: Add handling of EFI_FILE_SYSTEM_VOLUME_LABEL GetInfo() Pedro Falcato
2021-08-21 14:47 ` Pedro Falcato [this message]
2021-08-24 1:41 ` [edk2-platforms PATCH v2 0/5] Ext4Pkg: Fix bugs Michael D Kinney
2021-08-24 1:58 ` Michael D Kinney
2021-08-24 1:59 ` Michael D Kinney
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-list 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=20210821144711.39546-6-pedro.falcato@gmail.com \
--to=devel@edk2.groups.io \
/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