On 6/06/2006, at 12:46 AM, Tyson Whitehead wrote:
On Sun June 4 2006 22:07, Bob Tracy wrote:cards: 0 [ES1888 ]: ES1888 - ESS AudioDrive ES1888ESS AudioDrive ES1888 at 0x220, irq 5, dma1 1, dma2 5
On both my XP1000 and the PWS600au I have an ES1887 reported in /proc/ asound/cards. Inspection of the mother board of the two machines verifies that they are indeed ES1887s.
I have tried Tyson's patched Debian 2.6.14-2 kernel on the PWS600au and both the snd-es18xx and the snd-sb8 modules work fine. No problems with aplay or speaker-test and no problems breaking them with a ctrl-C.
Unfortunately I can't get that kernel to work on the XP1000 - I'm not using the onboard scsi and I specifically compile kernels such that it can't load the Qlogic scsi controller module (since the qlogic module barfs randomly and badly on this machine). The debian kernel loads the qlogic controller, presumably misnames all the scsi discs, and fails at the pivot_root. If I knew how it named the various discs I might be able to give a boot parameter to find the root drive.
I have also my self-compiled 2.6.16.18 kernel (standard linux sources rather than debian) installed on both computers. Both the patched snd-es18xx and the snd-sb8 modules fail under a ctrl-C of aplay. They produce the kernel bad page state messages. The PWS600au on one occasion even produced the kernel bad page state messages when aplay exited normally.
I haven't been able to try Tysons' newly patched 2.6.16 debian kernel. Nevertheless the evidence is mounting that something has happened since the 2.6.14 kernel to the 2.6.16 kernel that has rendered the alsa sound drivers on these alpha machines flakey.
I'm not going to be able to look further at this issue until the weekend.
Michael.