build error in ehci-hcd when disable PCI bus

DISCUSSION Yoshihiro Shimoda 4 participants 6 messages Feb 2, 2011 – Feb 3, 2011 SuperH

Share & tools

Yoshihiro Shimoda

author

2011-02-02 03:44 UTC #0 lore raw

Hi Paul,

I found the build error when I am debugging other board. Also I can duplicate this issue on sdk7786 when the PCI bus is disabled. I tested linux-2.6.38-rc3.
Do you think what souce code should I fix?

******* build Log *******
In file included from drivers/usb/host/ehci-hcd.c:340:
drivers/usb/host/ehci-sched.c: In function ‘ehci_quirk_amd_L1’:
drivers/usb/host/ehci-sched.c:1606: error: implicit declaration of function ‘outb_p’ drivers/usb/host/ehci-sched.c:1607: error: implicit declaration of function ‘inb_p’ drivers/usb/host/ehci-sched.c:1611: error: implicit declaration of function ‘outl_p’ drivers/usb/host/ehci-sched.c:1614: error: implicit declaration of function ‘inl_p’ make[3]: *** [drivers/usb/host/ehci-hcd.o] Error 1
**************************

Best regards,
Yoshihiro Shimoda

Nobuhiro Iwamatsu

2011-02-02 04:39 UTC #1 lore raw

Hi, Shimoda-san.

2011/2/2 Yoshihiro Shimoda <[email protected]>:

Quoted 7 lines

Hi Paul,
I found the build error when I am debugging other board. Also I can duplicate this issue on sdk7786 when the PCI bus is disabled. I tested linux-2.6.38-rc3.
Do you think what souce code should I fix?

Quoted 9 lines

******* build Log *******
In file included from drivers/usb/host/ehci-hcd.c:340:
drivers/usb/host/ehci-sched.c: In function ‘ehci_quirk_amd_L1’:
drivers/usb/host/ehci-sched.c:1606: error: implicit declaration of function ‘outb_p’ drivers/usb/host/ehci-sched.c:1607: error: implicit declaration of function ‘inb_p’ drivers/usb/host/ehci-sched.c:1611: error: implicit declaration of function ‘outl_p’ drivers/usb/host/ehci-sched.c:1614: error: implicit declaration of function ‘inl_p’ make[3]: *** [drivers/usb/host/ehci-hcd.o] Error 1
**************************

I am working to convert from in/out to __raw_read/__raw_write now in arch/sh .

Best regaerds,
Nobuhiro

Signature · 1 line

Nobuhiro Iwamatsu

Paul Mundt

2011-02-02 23:36 UTC #2 lore raw

In reply to Yoshihiro Shimoda

Quoted 5 lines · Yoshihiro Shimoda

I found the build error when I am debugging other board. Also I can duplicate this issue on sdk7786 when the PCI bus is disabled. I tested linux-2.6.38-rc3.
Do you think what souce code should I fix?

Quoted 10 lines

******* build Log *******
In file included from drivers/usb/host/ehci-hcd.c:340:
drivers/usb/host/ehci-sched.c: In function ?ehci_quirk_amd_L1?:
drivers/usb/host/ehci-sched.c:1606: error: implicit declaration of function ?outb_p? drivers/usb/host/ehci-sched.c:1607: error: implicit declaration of function ?inb_p? drivers/usb/host/ehci-sched.c:1611: error: implicit declaration of function ?outl_p? drivers/usb/host/ehci-sched.c:1614: error: implicit declaration of function ?inl_p? make[3]: *** [drivers/usb/host/ehci-hcd.o] Error 1
**************************

When the PCI bus is disabled we don't really have any concept of PIO, so PIO support and the associated accessors are explicitly disabled.

The quirk this refers to looks like it may depend on PCI, and if not, it certainly has a platform dependency that can be used to encapsulate it.

The proper fix here is to fix ehci-sched.c, not re-introduce broken PIO crap that wraps in to MMIO.

Nobuhiro Iwamatsu

2011-02-02 23:44 UTC #3 lore raw

In reply to Yoshihiro Shimoda

2011/2/3 Paul Mundt <[email protected]>:

Quoted 16 lines · Yoshihiro Shimoda

Quoted 5 lines
I found the build error when I am debugging other board. Also I can duplicate this issue on sdk7786 when the PCI bus is disabled. I tested linux-2.6.38-rc3.
Do you think what souce code should I fix?
Quoted 10 lines
******* build Log *******
In file included from drivers/usb/host/ehci-hcd.c:340:
drivers/usb/host/ehci-sched.c: In function ?ehci_quirk_amd_L1?:
drivers/usb/host/ehci-sched.c:1606: error: implicit declaration of function ?outb_p? drivers/usb/host/ehci-sched.c:1607: error: implicit declaration of function ?inb_p? drivers/usb/host/ehci-sched.c:1611: error: implicit declaration of function ?outl_p? drivers/usb/host/ehci-sched.c:1614: error: implicit declaration of function ?inl_p? make[3]: *** [drivers/usb/host/ehci-hcd.o] Error 1
**************************

Quoted 8 lines

When the PCI bus is disabled we don't really have any concept of PIO, so PIO support and the associated accessors are explicitly disabled.
The quirk this refers to looks like it may depend on PCI, and if not, it certainly has a platform dependency that can be used to encapsulate it.
The proper fix here is to fix ehci-sched.c, not re-introduce broken PIO crap that wraps in to MMIO.

Are these same to other devices?
Some device driver depends on PIO.

Signature · 1 line

Nobuhiro Iwamatsu

Nobuhiro Iwamatsu

2011-02-02 23:46 UTC #4 lore raw

In reply to Yoshihiro Shimoda

Sorry, I got advice from you in other threads.

2011/2/3 Nobuhiro Iwamatsu <[email protected]>:

Quoted 32 lines

2011/2/3 Paul Mundt <[email protected]>:
Quoted 16 lines · Yoshihiro Shimoda
Quoted 5 lines
I found the build error when I am debugging other board. Also I can duplicate this issue on sdk7786 when the PCI bus is disabled. I tested linux-2.6.38-rc3.
Do you think what souce code should I fix?
Quoted 10 lines
******* build Log *******
In file included from drivers/usb/host/ehci-hcd.c:340:
drivers/usb/host/ehci-sched.c: In function ?ehci_quirk_amd_L1?:
drivers/usb/host/ehci-sched.c:1606: error: implicit declaration of function ?outb_p? drivers/usb/host/ehci-sched.c:1607: error: implicit declaration of function ?inb_p? drivers/usb/host/ehci-sched.c:1611: error: implicit declaration of function ?outl_p? drivers/usb/host/ehci-sched.c:1614: error: implicit declaration of function ?inl_p? make[3]: *** [drivers/usb/host/ehci-hcd.o] Error 1
**************************
Quoted 8 lines
When the PCI bus is disabled we don't really have any concept of PIO, so PIO support and the associated accessors are explicitly disabled.
The quirk this refers to looks like it may depend on PCI, and if not, it certainly has a platform dependency that can be used to encapsulate it.
The proper fix here is to fix ehci-sched.c, not re-introduce broken PIO crap that wraps in to MMIO.
Are these same to other devices?
Some device driver depends on PIO.
--
Nobuhiro Iwamatsu

Signature · 3 lines

Nobuhiro Iwamatsu
   iwamatsu at {nigauri.org / debian.org}
   GPG ID: 40AD1FA6

Yoshihiro Shimoda

author

2011-02-03 01:29 UTC #5 lore raw

In reply to Yoshihiro Shimoda

Hi Paul-san,

Quoted 8 lines · Paul Mundt

When the PCI bus is disabled we don't really have any concept of PIO, so PIO support and the associated accessors are explicitly disabled.
The quirk this refers to looks like it may depend on PCI, and if not, it certainly has a platform dependency that can be used to encapsulate it.
The proper fix here is to fix ehci-sched.c, not re-introduce broken PIO crap that wraps in to MMIO.

Thank you very much for your reply.
I will report it to linux-usb ml.

Best regards,
Yoshihiro Shimoda