Someone Got Our Wifi Chip Working On A Bb

[Mithrildor]

And god said: "Let there be hope"
(if he wants to contribute , he is still active so that is already awesome)

http://greekclaud.blogspot.com/2009_08_01_archive.html

 

[mali]

He posted it on these boards, too

 

[Mithrildor]

/me feels stupid, link?

 

[mali]

/me feels stupid

Nah, this dedicated thread should generate more attention as it is missed easily, apparently

I don't know if these links are helpful or known yet:


http://greekclaud.blogspot.com/2009_08_01_archive.html
http://groups.google.com/group/beagleboard/browse_thread/thread/25fcddf411c5da57/ec89bda9a40c0d2c?lnk=raot
http://markmail.org/message/h3i55qgkxcfqejpo
http://permalink.gmane.org/gmane.comp.hardware.beagleboard.user/1952

In my case , on OMAP3530 the SDIO xfer fails in 2 cases.

1. source buffer is not in DMA:able space.
2. source buffer address is not 4 buf aligned.

BTW, I replaced both the omap_hsmmc and SDIO driver with drivers I ported from an omap 2430 implementation and solved all performance problems (4-bit mode with interrupts now working). This is on the 2.6.29-rc3 kernel with B7 beagleboard. I haven't looked yet but I'll bet this implementation uses pre- allocated kmalloced buffers as you suggested.

Using kmalloced DMA:able buffer before calling sdio_memcpy_fromio/sdio_memcpy_toio is a workaround.

drivers/mmc/host/omap.c switches to PIO when DMA is not possible due to constraints.

drivers/mmc/host/omap_hsmmc.c always attempts DMA. I tried host->use_dma = 0, but it didn't work.

compare definition of function mmc_omap_prepare_data in both the drivers.

Finially I get a workable WL1251 driver with beagleboard. The tcp performance ~10 Mbps with iperf. The DMA workaround is work like Kalpesh's hint. But the size of DMA need to adjust.

 

[Mithrildor]

I thought you meant claud posted here lol.

 

[mali]

[quote name"ED on German boards"]Und gleich mal an notaz / MWeston weitergeleitet. Klingt doch interessant
Forwarded it to notaz and MWeston. Sounds interesting [/quote]

 

[Mithrildor]

Hopefully he wants to contribute.

 

[Username]

This sounds promising.

 

[notaz]

That DMA problem prevents the driver from working completely, and it's only applicable to the old TI driver (Nokia one doesn't have this problem at all). Also that guy is using different wifi module (same chip but different manufacturer).

 

[Shock-Socks]

YAY! hopefully we'll be able to get things working with this, though i do feel bad for the pandora team cuz they just bought all of those mini wifi dongles. :\

 

[Asiyura]

Hope!

 

[Mithrildor]

That DMA problem prevents the driver from work completely, and it's only applicable to the old TI driver (Nokia one doesn't have this problem at all). Also that guy is using different wifi module (same chip but different manufacturer).

Is the difference in manufacturer a problem (as its the same chip). So this isnt going to fix our problem?

 

[skeezix]

Omap3530 chip. But different wifi module. Ie: you choose supplier ofmodule which is then packaged into your TI chip. A sub-chip sort of.

TI recommended the one in pandora :/

jeff

 

[xyta]

That DMA problem prevents the driver from work completely, and it's only applicable to the old TI driver (Nokia one doesn't have this problem at all). Also that guy is using different wifi module (same chip but different manufacturer).

It seems that Claud got wifi working at full speed. In his last setup ( http://greekclaud.blogspot.com/2009/08/wl1251-sdio-performance-and-setting.html ) the 1251 is connected via SDIO and Interrupt via GPIO to Bb. AFAIK that's exacly how the 1251 is connected to the OMAP in our Pandora?? The Muruta module seems to be just a little 1251 board with connectors for antenna and stuff...Pic
So I think we should ask him for his driver (I don't care if it is an old TI, a new Nokia, a Palm or whatever - driver if I can get full speed wifi)

Omap3530 chip. But different wifi module. Ie: you choose supplier ofmodule which is then packaged into your TI chip. A sub-chip sort of.

Um, really?

I can't really imagine that TI would produce several different versions of a chip with the same type number which are completely incompatible??

On the TI website
Link
there's only one TI1251 listed. And they offer the 1253 with a minor improved feature list ??????

How could that make sense for TI, offering several incompatible versions of the 1251 for use with their OMAPs? I can't understand that.

 

[Mithrildor]

Notaz says it wont work and is not needed.

 

[hobbyman II]

is not needed?, does this mean Notaz is onto something?, or they identified a hardware issue?, please tell, enquiring minds NEED TO KNOW....TELL ME.....TELL ME NOW OR THE KITTY GETS IT!!!!!!!!! :angry: h34r:



jk

 

[xyta]

Omap3530 chip. But different wifi module. Ie: you choose supplier ofmodule which is then packaged into your TI chip. A sub-chip sort of.

TI recommended the one in pandora :/

jeff

Perhaps I (or someone else interested) can do some more investigations. Which WL1251 subtype is used by our Pandora?

 

[mali]

I've answered on the German boards, too.
This one is used:
http://www.jorjin.com.tw/Product_SiPmodule_WG7210.htm

edit:
http://www.codejedi.com/blogmedia/entertainment/gaming/pandora/IMG_1037.JPG
http://gs113.photobucket.com/groups/n213/I0S2J7KION/?action=view&current=rev5.jpg

 

[notaz]

So I think we should ask him for his driver (I don't care if it is an old TI, a new Nokia, a Palm or whatever - driver if I can get full speed wifi)

Well according to him, he uses "WL1251 from linux wireless" with "Modification : none", we already have that. "WiLink 4.x from android source code" is also there.

Omap3530 chip. But different wifi module. Ie: you choose supplier ofmodule which is then packaged into your TI chip. A sub-chip sort of.

Um, really?

I can't really imagine that TI would produce several different versions of a chip with the same type number which are completely incompatible??

The module in pandora has various additional chips like BT, power management. Cloud seems to be using rather different configuration.

is not needed?, does this mean Notaz is onto something?

What I'm sure about is that it's not the DMA issue that has been linked to from first post of this thread.

 

[Exophase]

This datasheet might be useful to someone:

http://www.humyo.com/8102793/WiFiMod_WG7201-00-HDG-R15.pdf

It's for another module, but it uses the same wl1251 and power chipsets. I'm going to assume the peripheral interface goes to the wl1251 chip and is therefore the same as in the TI module, since nothing else makes an awful lot of sense. But I could be wrong, of course.

I was curious to see if SPI interface could be attempted instead of SDIO. This is how OMAP3530 muxes the two:

mmc2_clk -> mcspi3_clk
mmc2_cmd -> mcspi3_simo
mmc2_dat0 -> mcspi3_somi
mmc2_dat2 -> mcspi3_cs1
mmc2_dat3 -> mcspi3_cs0

Here is how the module muxes them, so presumably how the wl1251 always will:

sdio_clk -> spi_clk
sdio_cmd -> spi_din
sdio_d0 -> spi_dout
sdio_d3 -> spi_csx

So it looks like it matches.

There doesn't appear to be a means of configuring between SPI or SDIO electrically, so it must be done somehow over the CMD or DIN in software.

So has anyone tried to use the older SPI-based drivers? Of course the peak communication rate to the chip would be 1/4th, but..