I've got myself a hanky. (Was: SB 64 still not there)
Thanks all you guys for the help with the configuration of my SB AWE 64. I owe you all. I am some steps closer to a working solution now, but we're not there yet. First I rewrote the isapnp.conf file with all the suggestions incorporated. After the isapnp /etc/isapnp.conf initialisation I got the following... Board 1 has Identity 0c 0b a6 bd 16 c3 00 8c 0e: CTL00c3 Serial No 195476758 [checksum 0c] So that looks good (I hope). Then modprobe -a sound, and the speakers produced a short blip to, while lsmod showed the module was loaded with 24 pages (whatever that may mean). I did a cat /dev/sndstat and that also looked good Sound Driver:3.5.4-960630 (Thu May 13 18:06:48 CST 1999 root, Linux P200 2.0.36 #2 Sun May 9 20:53:11 CST 1999 i586 unknown) Kernel: Linux P200 2.0.36 #2 Thu May 13 18:08:47 CST 1999 i586 Config options: 0 Installed drivers: Type 1: OPL-2/OPL-3 FM Type 2: Sound Blaster Type 7: SB MPU-401 Card config: Sound Blaster at 0x220 irq 5 drq 1,5 SB MPU-401 at 0x330 irq 5 OPL-2/OPL-3 FM at 0x388 drq 0 Audio devices: 0: Sound Blaster 16 (4.16) Synth devices: 0: Yamaha OPL-3 Midi devices: 0: Sound Blaster 16 Timers: 0: System clock Mixers: 0: Sound Blaster Only the MPU-401 didn't have drq 0 as it should have. The big test came with this: cat chiquita.au > /dev/audio, however the following returned # cat chiquita.au > /dev/audio Sound error: Couldn't allocate DMA buffer bash: dev/audio: Cannot allocate memory Using /dev/dsp gave the same result. I thought it might have to do with the kernel compilation, so I did it over again according to the book, but no positive result. I am not really sure how much buffer I shoud assign. I have a AWE64 Value and that card has 512k RAM on board. When I compiled the kernel I entered 512k, so I hope that is right. If you have any ideas, then I hope you can pass them on. Thanks again. Background: I do a lot with plain audio, especially editing of sound files for my English classes. These files can run in the hundreds of MB and although Windblows95 can handle these without too much problems I need to run them over a LAN and that poses more problems. I'm not interested in MIDI, so if that doesn't work out it's okay. Cheers, Hans
Re: I've got myself a hanky. (Was: SB 64 still not there)
OKay, okay, that was of course HANKIE what I meant. - H
Re: I've got myself a hanky. (Was: SB 64 still not there)
On Thu, 13 May 1999, Hans van den Boogert wrote: Thanks all you guys for the help with the configuration of my SB AWE 64. I owe you all. I am some steps closer to a working solution now, but we're not there yet. [...] # cat chiquita.au /dev/audio Sound error: Couldn't allocate DMA buffer bash: dev/audio: Cannot allocate memory I'll bet you have more than 16MB of RAM, right? Okay, this is one of the annoying things about running a real operating system like Linux on historically bent hardware like the PC. It seems that DMA (direct memory access) on the ISA bus can only happen in the first 16MB of RAM. Linux likes to treat all of memory the same (because on sane hardware, it *is*) and by default doesn't reserve any space for DMA stuff below 16MB. This can cause problems if you want to allocate some of that memory and it's all filled up. (The floppy drive can suffer this problem, too.) There are a couple of solutions. Traditionally, you'd compile sound into the kernel, not as a module... but you and I have ISA PNP cards. I'm told that there's an option to the sound module, dmabuf=1, that forces it to grab a buffer and hold it as long as the module is loaded, but this may only be in the 2.2.x kernels. To make sure that this is really the problem, try two things. First, the next time you get that message, try using the floppy drive. Second, try booting the kernel with the option mem=16M. (If you're using LILO, try booting with linux mem=16M. This will force the kernel to use only memory below 16MB, and thus you should always have some memory free for DMA stuff, though you'll swap more. Sincerely, Ray Ingles (248) 377-7735 [EMAIL PROTECTED] Modern deductive method: 1) Devise hypothesis. 2) Apply for grant. 3) Perform experiments. 4) Revise hypothesis. 5) Backdate revised hypothesis. 6) Publish.
