[icecast-dev] Patches - Was: Stream metadata settings

Melanie melanie at t-data.com
Mon Jan 12 22:15:54 PST 2004



Hi,

On 2004.01.13 02:32 Michael Smith wrote:
> Since we've now (finally!) got 2.0 out the door, I'm now coming back to
> these
> patches - hoping to actually get the functionality incorporated.
> 
> I've got a few questions, though:
> 	1) Documentation: did you ever manage to write any? The options
> you added are
> fairly straightforward for the most part, so this shouldn't be too hard.

I put that off until I get some positive confirmation that the patches 
_will_ get incorporated. Naturally, I didn't want to waste my time 
documenting patches that would never make it into Icecast in the first 
place.
However, I'm not refusing to write it, expect some docs when I hear the 
patches are definitely going in.

> 	2) The 'no-mount' option: why? Does this have any effect at all
> that isn't
> handled by setting max-listeners to zero?

Fallback will add the clients from the upper level source to the fallback's 
pending queue. IIRC, that's where max-listeners gets checked. Setting 
max-listeners to zero would disallow any listeners at all. No-mount will 
only prevent mounting the source directly, by name, it will not prevent 
listening to it as a fallback. That was the idea.

> 	3) In the multilevel fallbacks patch, you've added a comment
> saying
> "Likely safe to do since there can only be one thread, ever", along with
> some
> static local variables, in the source_find_mount() function. I'm not sure
> why
> you said this, it's clearly incorrect: the source_* functions are called
> from
> many different threads, and this is clearly not safe. This is easily
> fixed,
> though, by passing the neccesary variables recursively, and splitting the
> function into two.

AFAIK, checking the function references I found that a certain lock must be 
hald to call it, off the top of my head I don't recall which. Since the 
function must be called under lock, there should only be one thread az a 
time executing it.
I wasn't 100% sure about this, that's why I added the comment, so someone 
else would have a look.

Melanie
--- >8 ----
List archives:  http://www.xiph.org/archives/
icecast project homepage: http://www.icecast.org/
To unsubscribe from this list, send a message to 'icecast-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body.  No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.



More information about the Icecast-dev mailing list