On Mon, Mar 28, 2011 at 06:09:38AM -0400, Mike Frysinger wrote: > On Mon, Mar 28, 2011 at 06:04, Felipe Balbi wrote: > > On Mon, Mar 28, 2011 at 05:57:21AM -0400, Mike Frysinger wrote: > >> On Mon, Mar 28, 2011 at 05:42, Felipe Balbi wrote: > >> > On Mon, Mar 28, 2011 at 05:26:46AM -0400, Mike Frysinger wrote: > >> >> >> any ideas on how to get this workaround merged ? another idea i had > >> >> >> was that we could add a hook to the top of dma_channel_program to let > >> >> >> targets muck with the arguments ... the Blackfin code would just need > >> >> >> to do something like this at the start of the func: > >> >> >> + if (musb_channel->transmit && channel->desired_mode == 1) > >> >> >> + len = len - (len % packet_sz); > >> >> > > >> >> > is that IP-related or Blackfin-only ? I mean, does this bug affects > >> >> > all > >> >> > MUSB of a particular RTL version or only the Blackfin ones ? > >> >> > >> >> our understanding of the issue is that it is Blackfin specific. seems > >> >> to be related to our internal bus/dma connections to the musb ip. > >> > > >> > ok. So we need a way to understand we need to apply that workaround only > >> > on blackfin. There we go with another blackfin-only driver flag... > >> > >> seems a shame to add a flag that will always evaluate to false for > >> non-Blackfin targets yet needs to be runtime checked for everyone ... > > > > I'm not keen in having ifdefs... Any other suggestions ? > > originally i was going to hijack the dma_channel_program pointer in > the Blackfin init, but dma_controller_create gets called too late for > that. > > so i was thinking a new func pointer in musb_platform_ops like > "pre_channel_program" which the normal dma_channel_program function > would check/call at the top. so i guess you might end up with the > same overhead for the most part (a logic test must occur), but at > least this way other people would be able to extend it if their own > needs called for it ?
Could be. MUSB has so many different bugs that most likely someone else might need this anyway. -- balbi _______________________________________________ Uclinux-dist-devel mailing list [email protected] https://blackfin.uclinux.org/mailman/listinfo/uclinux-dist-devel
