[rt2x00-users] Indirect register access failed

[Helmut Schaa]

Ping. Any update regarding this issue?

Am Sonntag 13 Juni 2010 schrieb Helmut Schaa:
> Am Freitag 11 Juni 2010 schrieb Antonio Quartulli:
> > On ven, giu 11, 2010 at 11:47:39 +0200, Gertjan van Wingerde wrote:
> > > On 06/11/10 11:41, Antonio Quartulli wrote:
> > > > On ven, giu 11, 2010 at 11:17:41 +0200, Helmut Schaa wrote:
> > > >> Am Freitag 11 Juni 2010 schrieb Antonio Quartulli:
> > > >>> On ven, giu 11, 2010 at 11:06:28 +0200, Helmut Schaa wrote:
> > > >>>> Am Freitag 11 Juni 2010 schrieb Antonio Quartulli:
> > > >>>>> On ven, giu 11, 2010 at 12:11:16 +0200, Vladimir Botka wrote:
> > > >>>>>> Jun 11 00:07:10 calpella kernel: [ 6828.735900] phy0 -> rt2800pci_load_firmware: Error - PBF system register not ready.
> > > >>>>>> Jun 11 00:07:10 calpella kernel: [ 6828.736466] phy0 -> rt2x00pci_regbusy_read: Error - Indirect register access failed: offset=0x00007010, value=0x55555555
> > > >>>>>>
> > > >>>>>
> > > >>>>> It seems the same error that happens to me!
> > > >>>>
> > > >>>> Yes, it's the exact same error :(
> > > >>>>
> > > >>>>> I got this error from kernel version 2.6.34....To use the driver I must
> > > >>>>> use the kernel 2.6.33 (with the latest driver from wireless-testing of
> > > >>>>> course!)
> > > >>>>
> > > >>>> So, you're using 2.6.33 + compat-wireless and it works? And when you switch
> > > >>>> to 2.6.34 + compat-wireless it fails?
> > > >>>>
> > > >>>
> > > >>> Exactly!
> > > >>> I did a lot of test with different version, and I always get the same
> > > >>> result: every woring driver on .33 has this problem on >33...should I
> > > >>> try .35-rc2?
> > > >>
> > > >> You could, but I guess that wouldn't give us more information about the
> > > >> issue itself.
> > > >>
> > > >> It's just strange, because if you use the same driver (rt2800pci from
> > > >> compat-wireless) the actual kernel version shouldn't matter much ...
> > > >>
> > > >> Maybe something more generic in the PCI code changed? But I'm not familiar
> > > >> enough with the PCI code to actually have an idea.
> > > > 
> > > > I am thinking the same, since in that part the driver doesn't call any
> > > > particular kernel function..it just reads!
> > > > 
> > > 
> > > What we may have to do is start analyzing the hardware on which this is going wrong
> > > (not just the WLAN NIC, but also the hardware platform itself).
> > > For instance, both on my netbook (Asus EEE-PC 1000H) and my PC/server (AMD dual-CPU)
> > > I am not experiencing this problem at all.
> > 
> > Actually I'm testing my NIC card on my SONY VAIO NR21Z, but, at this
> > point, I will test it on my EEEPC 901 and I will report the results.
> 
> Antonio, could you please check if the patch below helps? At least these
> register writes are done by the legacy drivers before the fw load.
> 
> Thanks,
> Helmut
> 
> PS: This patch is nor for being merged. Just for testing.
> 
> Signed-off-by: Helmut Schaa <helmut.schaa at googlemail.com>
> ---
>  drivers/net/wireless/rt2x00/rt2800pci.c |    4 ++++
>  1 files changed, 4 insertions(+), 0 deletions(-)
> 
> diff --git a/drivers/net/wireless/rt2x00/rt2800pci.c b/drivers/net/wireless/rt2x00/rt2800pci.c
> index 1a2079f..0ee92fa 100644
> --- a/drivers/net/wireless/rt2x00/rt2800pci.c
> +++ b/drivers/net/wireless/rt2x00/rt2800pci.c
> @@ -257,6 +257,10 @@ static int rt2800pci_load_firmware(struct rt2x00_dev *rt2x00dev,
>  	rt2x00_set_field32(&reg, WPDMA_GLO_CFG_TX_WRITEBACK_DONE, 1);
>  	rt2800_register_write(rt2x00dev, WPDMA_GLO_CFG, reg);
>  
> +	rt2800_register_write(rt2x00dev, WPDMA_RST_IDX, 0x1003f);
> +	rt2800_register_write(rt2x00dev, PBF_SYS_CTRL, 0xe1f);
> +	rt2800_register_write(rt2x00dev, PBF_SYS_CTRL, 0xe00);
> +
>  	/*
>  	 * enable Host program ram write selection
>  	 */
>