ALSA: bits vs bytes bug in snd_card_create()

PATCH In review Dan Carpenter 3 participants 6 messages Jan 23, 2014

Subsystems (2)

Sound Kernel janitors

lore

Share & tools

lore

Dan Carpenter

author

2014-01-23 08:21 UTC #0 lore raw

PATCH 1/1 ALSA: bits vs bytes bug in snd_card_create()

The test here is intended intended to prevent shift wrapping bugs when we do "1U << idx2". We should consider the number of bits in a u32 instead of the number of bytes.

Fixes 7bb2491b35a2 ('ALSA: Add kconfig to specify the max card numbers')

Signed-off-by Dan Carpenter <[email protected]>

---
WARNING: Please check this one carefully because I might be wrong.

sound/core/init.c +1 -1

index 1351f22f651c..e9e3db321e64 100644
@@ -183,7 +183,7 @@ int snd_card_create(int idx, const char *xid,
 	if (idx < 0) {
 		for (idx2 = 0; idx2 < SNDRV_CARDS; idx2++) {
 			/* idx == -1 == 0xffff means: take any free slot */
-			if (idx2 < sizeof(int) && !(idx & (1U << idx2)))
+			if (idx2 < 32 && !(idx & (1U << idx2)))
 				continue;
 			if (!test_bit(idx2, snd_cards_lock)) {
 				if (!slots[idx2] || !*slots[idx2]) {

Takashi Iwai

2014-01-23 08:36 UTC #1 lore raw

At Thu, 23 Jan 2014 11:21:28 +0300,

Quoted 9 lines · Dan Carpenter

The test here is intended intended to prevent shift wrapping bugs when we do "1U << idx2". We should consider the number of bits in a u32 instead of the number of bytes.
Fixes 7bb2491b35a2 ('ALSA: Add kconfig to specify the max card numbers')
Signed-off-by Dan Carpenter <[email protected]>
---
WARNING: Please check this one carefully because I might be wrong.

You're correct. I obviously didn't test more than 4. Applied (with Cc to stable) now, thanks.

Takashi

Quoted 15 lines

sound/core/init.c +1 -1

index 1351f22f651c..e9e3db321e64 100644
@@ -183,7 +183,7 @@ int snd_card_create(int idx, const char *xid,
 	if (idx < 0) {
 		for (idx2 = 0; idx2 < SNDRV_CARDS; idx2++) {
 			/* idx == -1 == 0xffff means: take any free slot */
-			if (idx2 < sizeof(int) && !(idx & (1U << idx2)))
+			if (idx2 < 32 && !(idx & (1U << idx2)))
 				continue;
 			if (!test_bit(idx2, snd_cards_lock)) {
 				if (!slots[idx2] || !*slots[idx2]) {

Takashi Iwai

2014-01-23 10:48 UTC #2 lore raw

At Thu, 23 Jan 2014 09:36:24 +0100,

Quoted 8 lines · Takashi Iwai

At Thu, 23 Jan 2014 11:21:28 +0300,
Quoted 5 lines · Dan Carpenter
The test here is intended intended to prevent shift wrapping bugs when we do "1U << idx2". We should consider the number of bits in a u32 instead of the number of bytes.

Quoted 7 lines

Quoted 4 lines
Fixes 7bb2491b35a2 ('ALSA: Add kconfig to specify the max card numbers')
Signed-off-by Dan Carpenter <[email protected]>
---
WARNING: Please check this one carefully because I might be wrong.
You're correct. I obviously didn't test more than 4. Applied (with Cc to stable) now, thanks.

I found there is another place doing it wrongly (a few lines above there), so fixed similarly in the same patch.

(And, yes, the code should be refactored; the patch will be sent soon
later.)

Takashi

-- 8< --
From: Dan Carpenter <[email protected]>
Subject: [PATCH] ALSA: bits vs bytes bug in snd_card_create()

The test here is intended intended to prevent shift wrapping bugs when we do "1U << idx2". We should consider the number of bits in a u32 instead of the number of bytes.

[fix another chunk similarly by tiwai]

Fixes 7bb2491b35a2 ('ALSA: Add kconfig to specify the max card numbers')

Signed-off-by Dan Carpenter <[email protected]>

Cc <[email protected]>

Signed-off-by Takashi Iwai <[email protected]>

---
 sound/core/init.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

sound/core/init.c +2 -2

index 1351f22f651c..e3c93cd77ee7 100644
@@ -170,7 +170,7 @@ int snd_card_create(int idx, const char *xid,
 	if (idx < 0) {
 		for (idx2 = 0; idx2 < SNDRV_CARDS; idx2++) {
 			/* idx == -1 == 0xffff means: take any free slot */
-			if (idx2 < sizeof(int) && !(idx & (1U << idx2)))
+			if (idx2 < 32 && !(idx & (1U << idx2)))
 				continue;
 			if (!test_bit(idx2, snd_cards_lock)) {
 				if (module_slot_match(module, idx2)) {

@@ -183,7 +183,7 @@ int snd_card_create(int idx, const char *xid,
 	if (idx < 0) {
 		for (idx2 = 0; idx2 < SNDRV_CARDS; idx2++) {
 			/* idx == -1 == 0xffff means: take any free slot */
-			if (idx2 < sizeof(int) && !(idx & (1U << idx2)))
+			if (idx2 < 32 && !(idx & (1U << idx2)))
 				continue;
 			if (!test_bit(idx2, snd_cards_lock)) {
 				if (!slots[idx2] || !*slots[idx2]) {

Signature · 1 line

1.8.5.2

walter harms

2014-01-23 08:55 UTC #3 lore raw

In reply to Dan Carpenter

Am 23.01.2014 09:21, schrieb Dan Carpenter:

Quoted 14 lines

The test here is intended intended to prevent shift wrapping bugs when we do "1U << idx2". We should consider the number of bits in a u32 instead of the number of bytes.
Fixes 7bb2491b35a2 ('ALSA: Add kconfig to specify the max card numbers')
Signed-off-by Dan Carpenter <[email protected]>
---
WARNING: Please check this one carefully because I might be wrong.
sound/core/init.c +0 -0
index 1351f22f651c..e9e3db321e64 100644

sound/core/init.c int snd_card_create(int idx, const char *xid,

Quoted 9 lines

sound/core/init.c +1 -2

@@ -183,7 +183,7 @@ int snd_card_create(int idx, const char *xid,
 	if (idx < 0) {
 		for (idx2 = 0; idx2 < SNDRV_CARDS; idx2++) {
 			/* idx == -1 == 0xffff means: take any free slot */
-			if (idx2 < sizeof(int) && !(idx & (1U << idx2)))
+			if (idx2 < 32 && !(idx & (1U << idx2)))
 				continue;
 			if (!test_bit(idx2, snd_cards_lock)) {
 				if (!slots[idx2] || !*slots[idx2]) {
--

Funny bug,
to be 100% sure you could multiply by CHAR_BIT ;)

other code simple does (1UL << (nr & 31))
on the other side a check like:

#if SNDRV_CARDS > 32

# error to much sound
#endif

whould give the user a hint what is going on instead of silently ignoring that error.

re,
wh

Quoted 4 lines

To unsubscribe from this list: send the line "unsubscribe kernel-janitors" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Takashi Iwai

2014-01-23 09:02 UTC #4 lore raw

At Thu, 23 Jan 2014 09:55:31 +0100,

Quoted 33 lines · walter harms

Am 23.01.2014 09:21, schrieb Dan Carpenter:
Quoted 14 lines
The test here is intended intended to prevent shift wrapping bugs when we do "1U << idx2". We should consider the number of bits in a u32 instead of the number of bytes.
Fixes 7bb2491b35a2 ('ALSA: Add kconfig to specify the max card numbers')
Signed-off-by Dan Carpenter <[email protected]>
---
WARNING: Please check this one carefully because I might be wrong.
sound/core/init.c +0 -0
index 1351f22f651c..e9e3db321e64 100644
sound/core/init.c int snd_card_create(int idx, const char *xid,
Quoted 9 lines
sound/core/init.c +1 -2
@@ -183,7 +183,7 @@ int snd_card_create(int idx, const char *xid,
 	if (idx < 0) {
 		for (idx2 = 0; idx2 < SNDRV_CARDS; idx2++) {
 			/* idx == -1 == 0xffff means: take any free slot */
-			if (idx2 < sizeof(int) && !(idx & (1U << idx2)))
+			if (idx2 < 32 && !(idx & (1U << idx2)))
 				continue;
 			if (!test_bit(idx2, snd_cards_lock)) {
 				if (!slots[idx2] || !*slots[idx2]) {
--
Funny bug,
to be 100% sure you could multiply by CHAR_BIT ;)
other code simple does (1UL << (nr & 31))
on the other side a check like:

Quoted 6 lines

#if SNDRV_CARDS > 32
# error to much sound
#endif
whould give the user a hint what is going on instead of silently ignoring that error.

SNDRV_CARDS can be more than 32. That's why the check was introduced.

Takashi

walter harms

2014-01-23 09:27 UTC #5 lore raw

Am 23.01.2014 10:02, schrieb Takashi Iwai:

Quoted 39 lines

At Thu, 23 Jan 2014 09:55:31 +0100,
Quoted 37 lines · walter harms
Am 23.01.2014 09:21, schrieb Dan Carpenter:
Quoted 14 lines
The test here is intended intended to prevent shift wrapping bugs when we do "1U << idx2". We should consider the number of bits in a u32 instead of the number of bytes.
Fixes 7bb2491b35a2 ('ALSA: Add kconfig to specify the max card numbers')
Signed-off-by Dan Carpenter <[email protected]>
---
WARNING: Please check this one carefully because I might be wrong.
sound/core/init.c +0 -0
index 1351f22f651c..e9e3db321e64 100644
sound/core/init.c int snd_card_create(int idx, const char *xid,
Quoted 9 lines
sound/core/init.c +1 -2
@@ -183,7 +183,7 @@ int snd_card_create(int idx, const char *xid,
 	if (idx < 0) {
 		for (idx2 = 0; idx2 < SNDRV_CARDS; idx2++) {
 			/* idx == -1 == 0xffff means: take any free slot */
-			if (idx2 < sizeof(int) && !(idx & (1U << idx2)))
+			if (idx2 < 32 && !(idx & (1U << idx2)))
 				continue;
 			if (!test_bit(idx2, snd_cards_lock)) {
 				if (!slots[idx2] || !*slots[idx2]) {
--
Funny bug,
to be 100% sure you could multiply by CHAR_BIT ;)
other code simple does (1UL << (nr & 31))
on the other side a check like:
#if SNDRV_CARDS > 32
# error to much sound
#endif

Quoted 6 lines

Quoted 2 lines
whould give the user a hint what is going on instead of silently ignoring that error.
SNDRV_CARDS can be more than 32. That's why the check was introduced.

yes of cause, but so far i understand this code will do "continue" if SNDRV_CARDS is more that 31, no warning.
(warning Start academic discussion):
If anyone ever has more that 31 soundcard he will wonder why 31 are working but the rest will remain silent. It would be nice to have a idea that there is an internal limit and it is not sufficient to change SNDRV_CARDS from 8 to 256. (did i understand the code correctly ?) and there is no warning.

re,
wh