Commenting out that line rather than playing with the constant seems to have worked wonders. What was the 1 second delay issue?
Rehan -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Maurizio Marini Sent: 29 June 2006 11:00 To: [email protected] Subject: Re: [Visdn-hackers] Re: Voice calls broken On Wed June 28 2006 05:24, stephen yong wrote: > In function hfc_sys_chan_write() in hfc-4s, I uncommented the line > 'printk(KERN_DEBUG "Trimming output fifo\n")' and > I got lots of these messages when I made a call. So I decided to disable > the next line "goto not_copying", and that fixed the problem! Maybe the > constant HFC_FIFO_JITTBUFF_HIGH (10) is set too low. The number of > used_octets varies from 11-383. Apparently this bit of code is not > found in the same function for the HFC-E1 driver. exactly! these changes were done after 607; we tried to change that piece of code playing around with HFC_FIFO_JITTBUFF_HIGH without success... reading Changelog: 2006-03-27 13:26:33 GMT Daniele Orlandi <[EMAIL PROTECTED]> patch-608 Summary: Attempt to trim output fifo, workaround for the 1-second delay issue Revision: isdn--devel--0.1--patch-608 -- Maurizio Marini GSM +39-335-8259739 Fano: +39-0721-855285 Milano +39-02303123406 S. Costanzo: +39-0721950396 IAXTel: (700) 350-1234 Crashing is the only thing windows does quickly. _______________________________________________ Visdn-hackers mailing list [email protected] https://mailman.uli.it/mailman/listinfo/visdn-hackers _______________________________________________ Visdn-hackers mailing list [email protected] https://mailman.uli.it/mailman/listinfo/visdn-hackers
