Skip to contentSkip to conversation

ext4fs: le32_to_cpu() used on a 16-bit field

RFC In review Rommel Custodio 7 participants 8 messages Jul 16, 2013 – Aug 5, 2013 1 message not in the archive ext4 file system

lore
Share & tools
lore
#0 lore raw

RFC 1/1 ext4fs: le32_to_cpu() used on a 16-bit field

Hi All,

U-Boot 2013.07-rc3 [ELDK 5.2.1 / ELDK 5.3]

Now I've started to use the new ext4 code. I need the "ext4write" command. Though there seems to be several problems with the ext2/ext4 code.

I am testing on an ml507 (PPC440, Big Endian).
There are some cases where the a field is 16-bit but le32_to_cpu() is used. Some checks (ie eh_magic) fails to match even if I use a correctly ext4 formatted MMC/SD card.

Does these seem right? Or am I mistaken?

-----

fs/ext4/ext4_common.c +0 -0
index 58880b4..22d4377 100644
@@ -1429,7 +1429,7 @@ static struct ext4_extent_header 

*ext4fs_get_extent_block
while (1) {
index = (struct ext4_extent_idx *)(ext_block + 1);

- if (le32_to_cpu(ext_block->eh_magic) != EXT4_EXT_MAGIC)
+ if (le16_to_cpu(ext_block->eh_magic) != EXT4_EXT_MAGIC)
return 0;

if (ext_block->eh_depth == 0)
@@ -1437,14 +1437,14 @@ static struct ext4_extent_header *ext4fs_get_extent_block
i = -1;
do {
i++;
- if (i >= le32_to_cpu(ext_block->eh_entries))
+ if (i >= le16_to_cpu(ext_block->eh_entries))
break;
} while (fileblock > le32_to_cpu(index[i].ei_block));

if (--i < 0)
return 0;

- block = le32_to_cpu(index[i].ei_leaf_hi);
+ block = le16_to_cpu(index[i].ei_leaf_hi);
block = (block << 32) + le32_to_cpu(index[i].ei_leaf_lo);

if (ext4fs_devread(block << log2_blksz, 0, fs->blksz, buf))
@@ -1543,17 +1543,17 @@ long int read_allocated_block(struct ext2_inode *inode, int fileblock)

do {
i++;
- if (i >= le32_to_cpu(ext_block->eh_entries))
+ if (i >= le16_to_cpu(ext_block->eh_entries))
break;
} while (fileblock >= le32_to_cpu(extent[i].ee_block)); if (--i >= 0) {
fileblock -= le32_to_cpu(extent[i].ee_block);
- if (fileblock >= le32_to_cpu(extent[i].ee_len)) {
+ if (fileblock >= le16_to_cpu(extent[i].ee_len)) {
free(buf);
return 0;
}

- start = le32_to_cpu(extent[i].ee_start_hi);
+ start = le16_to_cpu(extent[i].ee_start_hi);
start = (start << 32) +
le32_to_cpu(extent[i].ee_start_lo);
free(buf);
-----

(Sorry, I can't CC anyone directly as I'm using the gmane "post" interface)

All the best,
Rommel

#1 lore raw

Hi Rommel,

Quoted 7 lines

Hi All,

U-Boot 2013.07-rc3 [ELDK 5.2.1 / ELDK 5.3]

Now I've started to use the new ext4 code. I need the "ext4write" command. Though there seems to be several problems with the ext2/ext4 code.

The ext4 code at u-boot (especially ext4write) needs special attention.

It is on top of my list to refactor and clean up this code base (especially the ext4write command).

Quoted 7 lines

I am testing on an ml507 (PPC440, Big Endian).
There are some cases where the a field is 16-bit but le32_to_cpu() is used. Some checks (ie eh_magic) fails to match even if I use a correctly ext4 formatted MMC/SD card.

Does these seem right? Or am I mistaken?

I would need to double check it (since I'm using only ext4load at my target board currently).

Quoted 18 lines

-----

fs/ext4/ext4_common.c +1 -1
index 58880b4..22d4377 100644
@@ -1429,7 +1429,7 @@ static struct ext4_extent_header 
 *ext4fs_get_extent_block
 	while (1) {
 		index = (struct ext4_extent_idx *)(ext_block + 1);
 
-		if (le32_to_cpu(ext_block->eh_magic) !=
 EXT4_EXT_MAGIC)
+		if (le16_to_cpu(ext_block->eh_magic) !=
 EXT4_EXT_MAGIC) return 0;
fs/ext4/ext4_common.c +0 -0
 		if (ext_block->eh_depth == 0)

fs/ext4/ext4_common.c static struct ext4_extent_header · line 1441

Quoted 6 lines
fs/ext4/ext4_common.c +1 -1
@@ -1437,14 +1437,14 @@ static struct ext4_extent_header 
 *ext4fs_get_extent_block
 		i = -1;
 		do {
 			i++;
-			if (i >= le32_to_cpu(ext_block->eh_entries))
+			if (i >= le16_to_cpu(ext_block->eh_entries))

^^^^^ We had some problems with unaligned access to 16 bit values recently (when those start at 0x2). It was observed on our ARM Exynos4 based target.

Quoted 36 lines
fs/ext4/ext4_common.c +1 -1
 				break;
 		} while (fileblock > le32_to_cpu(index[i].ei_block));
 
 		if (--i < 0)
 			return 0;
 
-		block = le32_to_cpu(index[i].ei_leaf_hi);
+		block = le16_to_cpu(index[i].ei_leaf_hi);
 		block = (block << 32) +
 le32_to_cpu(index[i].ei_leaf_lo); 

if (ext4fs_devread(block << log2_blksz, 0,
fs->blksz, buf)) @@ -1543,17 +1543,17 @@ long int read_allocated_block(struct ext2_inode *inode, int fileblock)

fs/ext4/ext4_common.c +1 -1
 		do {
 			i++;
-			if (i >= le32_to_cpu(ext_block->eh_entries))
+			if (i >= le16_to_cpu(ext_block->eh_entries))
 				break;
 		} while (fileblock >=

le32_to_cpu(extent[i].ee_block)); if (--i >= 0) {

fs/ext4/ext4_common.c +0 -1
 			fileblock -= le32_to_cpu(extent[i].ee_block);
-			if (fileblock >=

le32_to_cpu(extent[i].ee_len)) {

fs/ext4/ext4_common.c +1 -0
+			if (fileblock >=

le16_to_cpu(extent[i].ee_len)) { free(buf);

fs/ext4/ext4_common.c +1 -2
 				return 0;
 			}
 
-			start = le32_to_cpu(extent[i].ee_start_hi);
+			start = le16_to_cpu(extent[i].ee_start_hi);
 			start = (start << 32) +
 					le32_to_cpu(extent[i].ee_start_lo);
 			free(buf);
-----
Quoted 2 lines

(Sorry, I can't CC anyone directly as I'm using the gmane "post" interface)

You can also CC ext4 u-boot's implementation authors.

Quoted 9 lines

All the best,
Rommel

_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Signature · 5 lines
Best regards,

Lukasz Majewski

Samsung R&D Institute Poland (SRPOL) | Linux Platform Group

A message that is not in the archive (sent off-list or not yet received): [email protected]

#3 lore raw

Sorry for the extra noise.

Quoted 39 lines · Rommel G Custodio

Dear Lukasz Majewski,

Quoted 6 lines · Lukasz Majewski

Hi Rommel,

Quoted 4 lines

Hi All,

U-Boot 2013.07-rc3 [ELDK 5.2.1 / ELDK 5.3]

Quoted 9 lines
Quoted 3 lines

Now I've started to use the new ext4 code. I need the "ext4write" command. Though there seems to be several problems with the ext2/ext4 code.

The ext4 code at u-boot (especially ext4write) needs special attention.

It is on top of my list to refactor and clean up this code base (especially the ext4write command).

I know the code is fairly recent and it seems limited testing (on a limited number of platforms) has only been performed.

Quoted 11 lines
Quoted 7 lines

I am testing on an ml507 (PPC440, Big Endian).
There are some cases where the a field is 16-bit but le32_to_cpu() is used. Some checks (ie eh_magic) fails to match even if I use a correctly ext4 formatted MMC/SD card.

Does these seem right? Or am I mistaken?

I would need to double check it (since I'm using only ext4load at my target board currently).

Without the patches below ext4load/ext2load doesn't work correctly on the ml507.

For testing, the default settings are used enabled). The SD card mounts correctly using ext4 in Linux but I get this error in u-boot:

Quoted 31 lines
ml507:/# ext4ls mmc 0:1
<DIR>       4096 .
<DIR>       4096 ..
<DIR>      16384 lost+found
         9294153 mybit.ace
ml507:/# ext2ls mmc 0:1
<DIR>       4096 .
<DIR>       4096 ..
<DIR>      16384 lost+found
        9294153 mybit.ace
ml507:/# ext4load mmc 0:1 1000000 /mybit.ace
invalid extent block
ml507:/# ext2load mmc 0:1 1000000 mybit.ace
invalid extent block
Quoted 16 lines

Review

Quoted 4 lines
            do {
                    i++;
(quoted diff) +1 -1
-                   if (i >= le32_to_cpu(ext_block->eh_entries))
+                   if (i >= le16_to_cpu(ext_block->eh_entries))

^^^^^ We had some problems with unaligned access to 16 bit values recently (when those start at 0x2). It was observed on our ARM Exynos4 based target.

Quoted 3 lines

(Sorry, I can't CC anyone directly as I'm using the gmane "post" interface)

You can also CC ext4 u-boot's implementation authors.

Forgot to Cc the list
And the original author's email seem to be bouncing.

Quoted 20 lines
Quoted 19 lines
Quoted 10 lines

All the best,
Rommel

_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

--
Best regards,

Lukasz Majewski

Samsung R&D Institute Poland (SRPOL) | Linux Platform Group

#4 lore raw
In reply to Rommel Custodio

On Tue, 16 Jul 2013 08:14:35 +0000 (UTC) Rommel Custodio sessyargc+uboot at gmail.com wrote,

Hi Rommel,

Quoted 8 lines

Hi All,

U-Boot 2013.07-rc3 [ELDK 5.2.1 / ELDK 5.3]

Now I've started to use the new ext4 code. I need the "ext4write" command. Though there seems to be several problems with the ext2/ext4 code.

Quoted 6 lines

I am testing on an ml507 (PPC440, Big Endian).
There are some cases where the a field is 16-bit but le32_to_cpu() is used. Some checks (ie eh_magic) fails to match even if I use a correctly ext4 formatted MMC/SD card.

Does these seem right? Or am I mistaken?

What kind of mailer program have you used to sent this e-mail? I receive following errors:

fatal: corrupt patch at line 111

I cannot apply this patch with either git am -3 or git apply.

Were you using git send-email?

Signature · 5 lines
Best regards,

Lukasz Majewski

Samsung R&D Institute Poland (SRPOL) | Linux Platform Group
#5 lore raw

PATCH RESEND 1/1 ext4fs: le32_to_cpu() used on a 16-bit field

Dear Lukasz Majewski

Quoted 3 lines · Lukasz Majewski

fatal: corrupt patch at line 111

I cannot apply this patch with either git am -3 or git apply.

I'm resending the patch.

Quoted 2 lines
Quoted 2 lines

(Sorry, I can't CC anyone directly as I'm using the gmane "post" interface)

All the best,
Rommel

---
 fs/ext4/ext4_common.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
fs/ext4/ext4_common.c +6 -6
index 2776293..ff9c4ec 100644
@@ -1432,7 +1432,7 @@ static struct ext4_extent_header *ext4fs_get_extent_block
 	while (1) {
 		index = (struct ext4_extent_idx *)(ext_block + 1);
 
-		if (le32_to_cpu(ext_block->eh_magic) != EXT4_EXT_MAGIC)
+		if (le16_to_cpu(ext_block->eh_magic) != EXT4_EXT_MAGIC)
 			return 0;
 
 		if (ext_block->eh_depth == 0)
@@ -1440,14 +1440,14 @@ static struct ext4_extent_header *ext4fs_get_extent_block
 		i = -1;
 		do {
 			i++;
-			if (i >= le32_to_cpu(ext_block->eh_entries))
+			if (i >= le16_to_cpu(ext_block->eh_entries))
 				break;
 		} while (fileblock > le32_to_cpu(index[i].ei_block));
 
 		if (--i < 0)
 			return 0;
 
-		block = le32_to_cpu(index[i].ei_leaf_hi);
+		block = le16_to_cpu(index[i].ei_leaf_hi);
 		block = (block << 32) + le32_to_cpu(index[i].ei_leaf_lo);
 
 		if (ext4fs_devread((lbaint_t)block << log2_blksz, 0, fs->blksz,
@@ -1548,17 +1548,17 @@ long int read_allocated_block(struct ext2_inode *inode, int fileblock)
 
 		do {
 			i++;
-			if (i >= le32_to_cpu(ext_block->eh_entries))
+			if (i >= le16_to_cpu(ext_block->eh_entries))
 				break;
 		} while (fileblock >= le32_to_cpu(extent[i].ee_block));
 		if (--i >= 0) {
 			fileblock -= le32_to_cpu(extent[i].ee_block);
-			if (fileblock >= le32_to_cpu(extent[i].ee_len)) {
+			if (fileblock >= le16_to_cpu(extent[i].ee_len)) {
 				free(buf);
 				return 0;
 			}
 
-			start = le32_to_cpu(extent[i].ee_start_hi);
+			start = le16_to_cpu(extent[i].ee_start_hi);
 			start = (start << 32) +
 					le32_to_cpu(extent[i].ee_start_lo);
 			free(buf);
Signature · 1 line
1.8.3.3
#6 lore raw
In reply to Lukasz Majewski

+Tom

Quoted 20 lines · Lukasz Majewski

On Tue, 16 Jul 2013 08:14:35 +0000 (UTC) Rommel Custodio sessyargc+uboot at gmail.com wrote,

Hi Rommel,

Quoted 8 lines

Hi All,

U-Boot 2013.07-rc3 [ELDK 5.2.1 / ELDK 5.3]

Now I've started to use the new ext4 code. I need the "ext4write" command. Though there seems to be several problems with the ext2/ext4 code.

Quoted 6 lines

I am testing on an ml507 (PPC440, Big Endian).
There are some cases where the a field is 16-bit but le32_to_cpu() is used. Some checks (ie eh_magic) fails to match even if I use a correctly ext4 formatted MMC/SD card.

Does these seem right? Or am I mistaken?

Quoted 9 lines

What kind of mailer program have you used to sent this e-mail? I receive following errors:

fatal: corrupt patch at line 111

I cannot apply this patch with either git am -3 or git apply.

Were you using git send-email?

It seems like we should try to fix this before the release?

Regards,
Simon

#7 lore raw

Hi all,

Quoted 27 lines · Simon Glass

+Tom

Quoted 23 lines · Lukasz Majewski

On Tue, 16 Jul 2013 08:14:35 +0000 (UTC) Rommel Custodio sessyargc+uboot at gmail.com wrote,

Hi Rommel,

Quoted 8 lines

Hi All,

U-Boot 2013.07-rc3 [ELDK 5.2.1 / ELDK 5.3]

Now I've started to use the new ext4 code. I need the "ext4write" command. Though there seems to be several problems with the ext2/ext4 code.

Quoted 6 lines

I am testing on an ml507 (PPC440, Big Endian).
There are some cases where the a field is 16-bit but le32_to_cpu() is used. Some checks (ie eh_magic) fails to match even if I use a correctly ext4 formatted MMC/SD card.

Does these seem right? Or am I mistaken?

What kind of mailer program have you used to sent this e-mail? I receive following errors:

Quoted 8 lines
Quoted 6 lines

fatal: corrupt patch at line 111

I cannot apply this patch with either git am -3 or git apply.

Were you using git send-email?

It seems like we should try to fix this before the release?

+1

I can confirm that reading ext4 extended header on BE machine (avr32) is broken. I'm currently on it to get it working by using Rommel's patch.

Regards

Andreas Bie?mann

Tested-by on v1 1/1
#8 lore raw
In reply to Rommel Custodio

Hi!

Quoted 11 lines

U-Boot 2013.07-rc3 [ELDK 5.2.1 / ELDK 5.3]

Now I've started to use the new ext4 code. I need the "ext4write" command. Though there seems to be several problems with the ext2/ext4 code.

I am testing on an ml507 (PPC440, Big Endian).
There are some cases where the a field is 16-bit but le32_to_cpu() is used. Some checks (ie eh_magic) fails to match even if I use a correctly ext4 formatted MMC/SD card.

Does these seem right? Or am I mistaken?

This fixed ext4 on powerpc-based board. Thanks!

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html