retry count?

retry count?

[Johannes Berg]

[-- Attachment #1: Type: text/plain, Size: 385 bytes --]

struct ieee80211_tx_status contains "retry_count" and I just noticed
that bcm43xx sets it to -1 if the frame was acked right away. Should it
be -1 in that case or should it be 0? A comment indicating either way
would be good. If it should be -1, then I suppose the radiotap tx status
needs to be fixed, it defines data_retries as a u8 so can't really be
less than 0.

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 190 bytes --]

Re: retry count?

[Johannes Berg]

[-- Attachment #1: Type: text/plain, Size: 572 bytes --]

On Sun, 2007-08-12 at 12:20 +0200, Johannes Berg wrote:
> struct ieee80211_tx_status contains "retry_count" and I just noticed
> that bcm43xx sets it to -1 if the frame was acked right away. Should it
> be -1 in that case or should it be 0? A comment indicating either way
> would be good. If it should be -1, then I suppose the radiotap tx status
> needs to be fixed, it defines data_retries as a u8 so can't really be
> less than 0.

Turns out that was wrong and b43 used to set it to -1 when the frame
wasn't transmitted at all due to some error.

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 190 bytes --]