fakephp problem in 2.6.22 (and 2.6.21)

[<john_flowers@agilent.com>]

Dear Linux PCI,

Kernel Version(s): 2.6.22 (Ubuntu 7.10) and 2.6.21 (Fedora Core 7)

I am trying to use fakephp to rescan the PCI bus when I reload my PCI Interface FPGA.

When I reload the FPGA, lspci shows "[disabled]" for the original Base Address Regions. My hope is that a rescan will re-enumerate the bus and fill in the PCI Settings on the FPGA interface. In fact, I can use setpci to re-enable the board after loading FPGA, but only if I have not changed any of the core parameters. I would hope that a rescan would read the possibly changed core parameters and reallocate memory as needed.

fakephp is loaded in to the kernel using "modprobe fakephp debug=1" as advised in order to get debug information in dmesg.

dmesg indicates fakephp is installed

fakephp: Fake PCI Hot Plug Controller Driver
fakephp: slot->name = 0000:00:00.0
...
fakephp: slot->name = 0000:03:00.0
fakephp: slot->name = 0000:06:03.0

my FPGA board is connected via PCI Express to bus 3 so I issue the following command

echo 1 > /sys/bus/pci/slots/0000\:03\:00.0/power
-bash: echo: write error: No such device

lspci still shows that the BAR on 03:0.0 are still disabled, and dmesg gives no other indications.

writing a 1 to any other device/slot seems to have no effect.

writing a 0 to this power file causes the terminal to hang, but something does happen in dmesg.

echo 0 > /sys/bus/pci/slots/0000\:03\:00.0/power ***TERMINAL HANG***

(open another terminal)
last few lines of dmesg 
fakephp: disable_slot - physical_slot = 0000:03:00.0
fakephp: No device in slot found
fakephp: No device in slot found
fakephp: No device in slot found
fakephp: No device in slot found
fakephp: No device in slot found
fakephp: No device in slot found
fakephp: No device in slot found
fakephp: removing slot 0000:03:00.0

Now echo 1 > /sys/bus/pci/slots/0000\:03\:00.0/power ***TERMINAL HANG***

(open another terminal)
no change in dmesg
echo 1 > /sys/bus/pci/slots/0000\:06\:03.0/power
-bash: echo: write error: No such device

now dmesg has a lot of detail:

fakephp: New device on PCI Bus #03 function 0:0
fakephp: slot->name = 0000:03:00.0
kobject_add failed for 0000:03:00.0 with -EEXIST, don't try to register things with the same name in the same directory.
 [<c04e376b>] kobject_shadow_add+0x153/0x185
 [<c04e3873>] kobject_register+0x19/0x30
 [<c04f48b5>] pci_hp_register+0x88/0x294
 [<f8d9921d>] add_slot+0xb3/0x10c [fakephp]
 [<f8d99377>] pci_rescan_buses+0x101/0x222 [fakephp]
 [<f8d9947c>] pci_rescan_buses+0x206/0x222 [fakephp]
 [<f8d994a2>] enable_slot+0xa/0x10 [fakephp]
 [<c04f4b55>] power_write_file+0x94/0xcd
 [<c04f4ac1>] power_write_file+0x0/0xcd
 [<c04f4042>] hotplug_slot_attr_store+0x1e/0x22
 [<c04abb8e>] sysfs_write_file+0xbd/0xe7
 [<c04abad1>] sysfs_write_file+0x0/0xe7
 [<c0475d33>] vfs_write+0xa8/0x154
 [<c0476342>] sys_write+0x41/0x67
 [<c0404f70>] syscall_call+0x7/0xb
 =======================
pci_hotplug: Unable to register kobject<3>fakephp: pci_hp_register failed with error -22

lspci still shows that my FPGA board BAR is "[disabled]"

Any advice would be appreciated.

John
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/