Skip to contentSkip to conversation

Linux v2.6.18-rc3

DISCUSSION Hubert Tonneau 2 participants 2 messages Aug 2, 2006

lore
Share & tools
lore
#0 lore raw
Quoted 4 lines · Takashi Iwai

Do you really get Oops with 2.6.17-final, too?
AFAIK, there was a bug in the early 2.6.17-rc regarding OSS emulation, but it got fixed in the later rc.

I confirm that the only oops I got was with 2.6.17-rc1

Quoted 3 lines

The behavior of invalid O_* flag seems incompatible with the older version. Although I don't think it's totally wrong behavior, surely better to keep the compatibility. The patch below should fix it.

I don't care: I have changed my player.
I also tend to agree that opening /dev/dsp read write was not a good idea, so I would not criticise the new behaviour.

Each version of Linux 2.6 tend to break some softwares due to subtil changes. What is bad is that these changes tend to be not announced, but on the other hand the kernel mailing list is very responsive, so all in all it might be a reasonable way to go.

#1 lore raw

At Wed, 02 Aug 2006 18:35:26 GMT,

Quoted 9 lines · Hubert Tonneau
Quoted 4 lines · Takashi Iwai

Do you really get Oops with 2.6.17-final, too?
AFAIK, there was a bug in the early 2.6.17-rc regarding OSS emulation, but it got fixed in the later rc.

I confirm that the only oops I got was with 2.6.17-rc1

Quoted 7 lines
Quoted 3 lines

The behavior of invalid O_* flag seems incompatible with the older version. Although I don't think it's totally wrong behavior, surely better to keep the compatibility. The patch below should fix it.

I don't care: I have changed my player.
I also tend to agree that opening /dev/dsp read write was not a good idea, so I would not criticise the new behaviour.

Well, incompatibility is worse in most cases than rationality. Could you test the patch really fixes your case, so that I can push it to 2.6.17-stable tree?

Quoted 4 lines

Each version of Linux 2.6 tend to break some softwares due to subtil changes. What is bad is that these changes tend to be not announced, but on the other hand the kernel mailing list is very responsive, so all in all it might be a reasonable way to go.

It was an unintentioal change, so it's not somthing to be announced ;) It's a thing just to be fixed.

Takashi