[U-Boot-Users] PCI OHCI / testing-USB branch
Hi,
is someone currently working on the testing-USB branch? Does it contain the most up to date USB code? I have to add support for a PCI OHCI controller on a 405 board. This will result in some modifications (endianess + bus offsets) of drivers/usb_ohci.c.
Is someone else using a PCI OHCI controller on a board that I might have missed to check?
Matthias
Dear Matthias,
Matthias Fuchs matthias.fuchs@esd-electronics.com writes:
is someone currently working on the testing-USB branch? Does it contain the most up to date USB code?
Yes, and no. I've got a couple of patches for the generic OHCI driver which have not been merged yet, because I want to do some more testing. I'll try to do this soon.
I have to add support for a PCI OHCI controller on a 405 board. This will result in some modifications (endianess + bus offsets) of drivers/usb_ohci.c.
Ok.
Is someone else using a PCI OHCI controller on a board that I might have missed to check?
Not that I know of.
Regards
Markus Klotzbuecher
Dear Markus,
is someone currently working on the testing-USB branch? Does it
contain
the most up to date USB code?
Yes, and no. I've got a couple of patches for the generic OHCI driver which have not been merged yet, because I want to do some more testing. I'll try to do this soon.
I was working on that too, but I only started today, so I'm glad I'm reading this posting now instead of two weeks later ;-). Could you please send me your patches, or even post them here? I could help you debugging it...
I have to add support for a PCI OHCI controller on a 405 board. This
will
result in some modifications (endianess + bus offsets) of drivers/usb_ohci.c.
I'm currently working on the NXP (previously Philips Semiconductors) PNX8535 SoC with a MIPS 4Kec core in it. We're running it in little endian mode, but I'm to keep everything big endian compatible using readl/writel from io.h.
Is someone else using a PCI OHCI controller on a board that I might
have
missed to check?
Yes, I am, but we haven't pushed back our contributions yet. We're still working on U-Boot 1.1.3, while at the same time I'm porting 1.1.4. Unfortunately 1.1.5 came before I finished and scr*wed everything up again. I could send you my work, but I doubt if one day of work on a 1.1.3 file would be of any help ;-)
With kind regards,
Robert.
Dear Robert,
Robert Delien robert.delien@nxp.com writes:
is someone currently working on the testing-USB branch? Does it
contain
the most up to date USB code?
Yes, and no. I've got a couple of patches for the generic OHCI
driver
which have not been merged yet, because I want to do some more testing. I'll try to do this soon.
I was working on that too, but I only started today, so I'm glad I'm reading this posting now instead of two weeks later ;-). Could you please send me your patches, or even post them here? I could help you debugging it...
It's now all in the testing-USB branch. Please also see doc/README.generic_usb_ohci for a brief description.
Regards
Markus Klotzbuecher
Hi Markus,
do you see a problem in passing a pointer to the gohci structure to usb_cpu_init() or at least to usb_board_init()?
In order to get the base address of my PCI OHCI controller I first have to find the device on the PCI bus. I think usb_board_init() is the right place for this. But I need &gohci to pass it back to the usb driver.
Here's is what I mean:
diff --git a/drivers/usb_ohci.c b/drivers/usb_ohci.c index c5e4c38..8afadd1 100644 --- a/drivers/usb_ohci.c +++ b/drivers/usb_ohci.c @@ -1680,19 +1680,20 @@ static char ohci_inited = 0;
int usb_lowlevel_init(void) { + memset (&gohci, 0, sizeof (ohci_t));
#ifdef CFG_USB_OHCI_CPU_INIT /* cpu dependant init */ - if(usb_cpu_init()) + if(usb_cpu_init(&gohci)) return -1; #endif
#ifdef CFG_USB_OHCI_BOARD_INIT /* board dependant init */ - if(usb_board_init()) + if(usb_board_init(&gohci)) return -1; #endif - memset (&gohci, 0, sizeof (ohci_t)); + memset (&urb_priv, 0, sizeof (urb_priv_t));
/* align the storage */ @@ -1719,8 +1720,10 @@ int usb_lowlevel_init(void) gohci.disabled = 1; gohci.sleeping = 0; gohci.irq = -1; - gohci.regs = (struct ohci_regs *)CFG_USB_OHCI_REGS_BASE;
+#ifdef CFG_USB_OHCI_REGS_BASE + gohci.regs = (struct ohci_regs *)CFG_USB_OHCI_REGS_BASE; +#endif gohci.flags = 0; gohci.slot_name = CFG_USB_OHCI_SLOT_NAME;
Regards Matthias
On Tuesday 28 November 2006 14:24, Markus Klotzbücher wrote:
Dear Robert,
Robert Delien robert.delien@nxp.com writes:
is someone currently working on the testing-USB branch? Does it
contain
the most up to date USB code?
Yes, and no. I've got a couple of patches for the generic OHCI
driver
which have not been merged yet, because I want to do some more testing. I'll try to do this soon.
I was working on that too, but I only started today, so I'm glad I'm reading this posting now instead of two weeks later ;-). Could you please send me your patches, or even post them here? I could help you debugging it...
It's now all in the testing-USB branch. Please also see doc/README.generic_usb_ohci for a brief description.
Regards
Markus Klotzbuecher
Hi Matthias,
Matthias Fuchs matthias.fuchs@esd-electronics.com writes:
do you see a problem in passing a pointer to the gohci structure to usb_cpu_init() or at least to usb_board_init()?
In order to get the base address of my PCI OHCI controller I first have to find the device on the PCI bus. I think usb_board_init() is the right place for this. But I need &gohci to pass it back to the usb driver.
Thats ok with me. I think it makes sense to pass the pointer to both board and cpu init. But please provide a proper patch.
Thanks!
Markus Klotzbücher
Hi
I am stuck at the position where iam not getting any data on USB bus. I am using the ISP1562 chip on my PCI and using the OHCI code. I am able to reset and initialize the controller as i am able to see the reset signal and High speed chirp signal on the bus but there is no data on the bus. It shows the invalid packet on with all 0xFF at the first time i connect the mass storage device and running the usb reset command what could be the problem ? How i have to map the PCI memory ? I though that it could be the invalid memory access as i am taking the gohci and hcca structure by just taking it as an array and the controller HcHcca registers are unable to access that location it 0x7ffa0000 in my case where my PCI memory is mapped at 0x80000000
Need some help Regards
Markus Klotzbücher-2 wrote:
Dear Matthias,
Matthias Fuchs matthias.fuchs@esd-electronics.com writes:
is someone currently working on the testing-USB branch? Does it contain the most up to date USB code?
Yes, and no. I've got a couple of patches for the generic OHCI driver which have not been merged yet, because I want to do some more testing. I'll try to do this soon.
I have to add support for a PCI OHCI controller on a 405 board. This will result in some modifications (endianess + bus offsets) of drivers/usb_ohci.c.
Ok.
Is someone else using a PCI OHCI controller on a board that I might have missed to check?
Not that I know of.
Regards
Markus Klotzbuecher
Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=D... _______________________________________________ U-Boot-Users mailing list U-Boot-Users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/u-boot-users
participants (4)
-
Markus Klotzbücher -
Matthias Fuchs -
reckless -
Robert Delien