Hello,
01.02.2012 6:18, Loh Tien Hock:
Hi Nikolai,
Sorry for the late reply.
Yes, ozlabs patchwork code is refactored from synopsys version, as far
as I understand. Are you referring to the parameter OTG_EN_DED_TX_FIFO
Well, this exact name does not seem to appear in either sysnopsys nor
apm version, so maybe you are referring to some internal or intermediate
version. The name looks like it is relevant though.
in the configuration? If that's the case, the core in our platform
does uses dedicated FIFO.
That quite explains why the driver works for you.
In sysnopsys version there is a en_multiple_tx_fifo module parameter
(labeled "Dedicated Non Periodic Tx FIFOs 0=disabled 1=enabled"), so you
could force it to 0 even if your core would normally allow (and default
to) 1. Having such parameter is exellent for testing/debugging BTW. Now,
in apm version (ozlabs patchwork) this module parameter has somehow
vanished, but searching for "en_multiple_tx_fifo" in the code would
still give you sufficient input for some testing without the use of
dedicated FIFOs (that is, if you were going to make sure the driver is
not broken, but do not have all various sorts of hardware implementation
by hand).
Thank you.
Nikolai
Thanks.
On Wed, Jan 25, 2012 at 8:26 PM, Nikolai Zhubr<n-a-zhubr@xxxxxxxxx> wrote:
Hello,
24.01.2012 20:02, Loh Tien Hock:
Hi Nikolai.
I tried the patch with g_serial. it needed some fix I have in my
company's repository for slave mode. Dma mode works correctly.
Because the code in ozlabs patchwork does not seem to have anything new or different (compared to synopsys version) with respect to the problem I'm observing, I'd guess you only tried core(s) with a so called "Dedicated FIFO Enhancement". Then it might work indeed.
However, the driver should either refuse to work without dedicated FIFOs present and state so honestly, or actually support such configuration correctly. Otherwise, it is still broken.
Personally, I'd prefer this configuration supported rather then refused, of course.
Thank you.
Nikolai
g_ether initializes correctly in kernel 3.0. There's a fix in g_ether
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
