Re: Is Familiar dead?

From: Paul Sokolovsky <pmiscml_at_gmail.com>
Date: Wed, 19 Dec 2007 03:49:48 +0200

Hello sza2,

Tuesday, December 18, 2007, 12:09:44 PM, you wrote:

> Hi Paul,

> It works! With the latest cvs source I can use the it again.

  Thanks for testing! I'm tagging -hh20 right now.

> I also
> have a SanDisk CF, which have ID 0x0045/0x0401. The Kingston card is
> detected a 0x0000/0x0000 as before so it seems it is really 0:0.

  So, if you have specific patches, please post them. As for Kingstom,
look at 2 (actually, more) different formats ide_cs.c has:

        PCMCIA_DEVICE_MANF_CARD(0x0045, 0x0401), /* SanDisk CFA */
        PCMCIA_DEVICE_PROD_ID12("ARGOSY", "CD-ROM", 0x78f308dc, 0x66536591),

  You'll likely need to use second format. But what those scary
numbers at the end mean is up to you to figure out, or use IRC, as I
suggested ;-).

> /sza2
>
>
> Paul Sokolovsky <pmiscml_at_gmail.com> írta:
>
> Hello Tamas,
>
> Monday, December 17, 2007, 11:38:01 PM, you wrote:
>
>> Hi Paul,
>
>> As the current kernel not working, I tried with my good old -hh14.
>
>> pccardctl info:
>
>> PRODID_1="SAMSUNG"
>> PRODID_2="04/05/06"
>> MANFID=0000,0000
>> FUNCID=4
>
> Hm, this apparently too little (well, fairly speaking, I dunno much
> details about adding new card support, I'll likely suggest you go
> catching someone on IRC who knows; hrw on #oe might be helpful).
>
>> Could you please tell me bit more detailed how can I test CVS HEAD?
>
> This seems like hard question in our context ;-I. You build kernel
> yourself, right? So, just do "cvs -z9 update -dP -A", that will
> unconditionally get you to HEAD, and build as usual.
>
>> /sza2
>
>> Paul Sokolovsky wrote:
>>> Hello Tamas,
>>>
>>> Monday, December 17, 2007, 10:32:39 PM, you wrote:
>>>
>>>> Hello Paul,
>>>
>>>> Well, the current problem seem to be the one you say. I installed latest
>>>> Angstrom distro with the -hh19 kernel and the CF is not working.
>>>
>>>> From dmesg:
>>>
>>>> egpio clear:reg 0 = 0x0001
>>>> egpio clear:reg 0 = 0x0001
>>>> egpio clear:reg 0 = 0x0001
>>>> cs: pcmcia_socket0: time out after reset.
>>>> egpio clear:reg 0 = 0x0001
>>>
>>>> Is there any remedy for this?
>>>
>>> This is tracked at http://bugs.openembedded.org/show_bug.cgi?id=3484
>>> . Philipp Zabel has just committed a supposed fix for that, but
>>> neither he nor me can test it immediately due to lack of a CF card on
>>> hands. So, if you can test CVS HEAD and report to the bugtracker, that
>>> would be appreciated.
>>>
>>>> Regarding the card id, I have Kingston Elite Pro CFs and it seems the
>>>> id is 0x0000, 0x0000. Isn't it so strange?
>>>
>>> Yes, that's rather strange. But there seem to be different formats
>>> of card IDs in the kernel. What pccardctl info/lspcmcia say for it and
>>> what /sys has for it?
>>>
>>>
>>>> /sza2
>>>
>>> []
>>>
>>>
>>>
>
>
>
> --
> Best regards,
> Paul pmiscml_at_gmail.com
>
>

> Ajándékvásárlás az utolsó pillanatban? A megoldás az Alexandra
> Online áruház. Még a december 21-én leadott könyv, CD, DVD
> megrendeléseket is 1 napos kiszállítással teljesítjük.
> http://ad.adverticum.net/b/cl,1,6022,236565,281483/click.prm

-- 
Best regards,
 Paul                            mailto:pmiscml_at_gmail.com
Received on Tue Dec 18 2007 - 20:44:28 EST

This archive was generated by hypermail 2.2.0 : Tue Dec 18 2007 - 20:45:05 EST