Bug 180997

Created attachment 124556 [details]
Output from lspci, lsmod and cat'ing files from /proc/asound

This files contains output for lspci lsmod and output from files in
/proc/asound when booting under kernel-xen-hypervisor-2.6.15-1.51_FC5

Created attachment 124557 [details]
Output from lspci, lsmod, and /proc/asound

This files contains output for lspci lsmod and output from files in
/proc/asound when booting under kernel-2.6.15-1.1928_FC5

These bugs are being closed since a large number of updates have been released
after the FC5 test1 and test2 releases. Kindly update your system by running yum
update as root user or try out the third and final test version of FC5 being
released in a short while and verify if the bugs are still present on the system
.Reopen or file new bug reports as appropriate after confirming the presence of
this issue. Thanks

FC5T3 is already available here: 
ftp://ftp.dti.ad.jp/pub/Linux/Fedora/core/test/4.92/  I have downloaded and 
done a clean install and still suffer from the same problem with kernel-xen-
hypervisor-2.6.15-1.1955_FC5.  Sound functions with kernel-2.6.15-1.1955_FC5.

I have the following line in /var/log/dmesg when boot with kernel-xen-
hypervisor:
PCI: Found IRQ 5 for device 0000:02:03.0

But /proc/interrupts reads:
           CPU0
  1:         10        Phys-irq  i8042
  7:          4        Phys-irq  parport0
  8:          1        Phys-irq  rtc
  9:          0        Phys-irq  acpi
 10:      14162        Phys-irq  yenta, yenta, uhci_hcd:usb1, ohci1394, peth0
 12:       2336        Phys-irq  i8042
 14:      26839        Phys-irq  ide0
256:     280954     Dynamic-irq  timer0
257:          0     Dynamic-irq  resched0
258:          0     Dynamic-irq  callfunc0
259:         85     Dynamic-irq  xenbus
260:          0     Dynamic-irq  console
NMI:          0
LOC:          0
ERR:          0
MIS:          0

/proc/interrupts in native kernel, for comparison, reads:
           CPU0
  0:   11987784          XT-PIC  timer
  1:         12          XT-PIC  i8042
  2:          0          XT-PIC  cascade
  5:          4          XT-PIC  Maestro3
  7:          4          XT-PIC  parport0
  8:          1          XT-PIC  rtc
  9:          1          XT-PIC  acpi
 10:   15662864          XT-PIC  yenta, yenta, ohci1394, uhci_hcd:usb1, eth0
 12:       2336          XT-PIC  i8042
 14:     435778          XT-PIC  ide0
NMI:          0
LOC:          0
ERR:          0
MIS:          0

I found the cause, and a work around that seems to do the job very well for me, 
however I am completely unsure of the negative implications of what I've done.  

I'm sure someones computer would get up and make a trip to taco bell in the 
early AM if they did this on their computer.  This is a definate case of a 
little bit of knowledge is a very dangerous thing.

I noticed the message in /var/log/dmesg that read:
"architecture does not support 28bit PCI busmaster DMA"
when booting with the xen kernel.

I downloaded and installed the kernel source RPM and tracked the message down  
to /usr/src/linux/sound/pci/maestro3.c

I ended up changing the lines that read:
        /* check, if we can restrict PCI DMA transfers to 28 bits */
        if (pci_set_dma_mask(pci, 0x0fffffff) < 0 ||
            pci_set_consistent_dma_mask(pci, 0x0fffffff) < 0) {
                snd_printk(KERN_ERR "architecture does not support 28bit PCI 
busmaster DMA\n");
                pci_disable_device(pci);
                return -ENXIO;
        }

to read like this (commented out the printk, pci disable and return):
        /* check, if we can restrict PCI DMA transfers to 28 bits */
        if (pci_set_dma_mask(pci, 0x0fffffff) < 0 ||
            pci_set_consistent_dma_mask(pci, 0x0fffffff) < 0) {
                //snd_printk(KERN_ERR "architecture does not support 28bit PCI 
busmaster DMA\n");
                //pci_disable_device(pci);
                //return -ENXIO;
        }

From there I compiled the module and replaced the existing snd-maestro3.ko 
in /lib/modules/2.6.15-1.2032_FC5xen0/kernel/sound/pci and rebooted to reload 
all the sound modules and now I have good working sound, with no obvious ill 
effects.

Device only support 28bit addresses (i.e. <=256MB RAM)
With your workaround, if you have less than 256MB RAM, and while you are lucky
and your buffers happen to be in the low 256MB, it will work, otherwise, will fail.

Trying to find anyways, why xen is not support 28bits DMA.

Later, Juan.

Please post your full dmesg from bootup.

Created attachment 126258 [details]
Copy of /var/log/dmesg

Created attachment 126259 [details]
Output from 'xm dmesg'

Please confirm this BZ is still current with the latest FC5 updates.

Fedora Core 5 and Fedora Core 6 are, as we're sure you've noticed, no longer
test releases. We're cleaning up the bug database and making sure important bug
reports filed against these test releases don't get lost. It would be helpful if
you could test this issue with a released version of Fedora or with the latest
development / test release. Thanks for your help and for your patience.

[This is a bulk message for all open FC5/FC6 test release bugs. I'm adding
myself to the CC list for each bug, so I'll see any comments you make after this
and do my best to make sure every issue gets proper attention.]

change QA contact

This report targets FC5, which is now end-of-life.

Please re-test against Fedora 7 or later, and if the issue persists, open a new bug.

Thanks