[ECOS] AT91 USB driver

[ECOS] AT91 USB driver

[Gábor Török]

Hi!

Have anybody tried the AT91 USB driver with other controllers than the
AT91SAM7S (I only tried with this)? I would like to use it for the
AT91SAM7SE, and after a quick compsrison between the USB registers of
the two chips, I noticed no major differences. The only thing was,
that the SAM7SE has more endpoints and the pullup is integrated, but
all the register and bit definitions are the same (except the bit
controlling the pullup and the registers of the extra endpoints).

Gábor

--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

Re: [ECOS] AT91 USB driver

[Frank Pagliughi]

Gábor Török wrote:
> Hi!
>
> Have anybody tried the AT91 USB driver with other controllers than the
> AT91SAM7S (I only tried with this)? I would like to use it for the
> AT91SAM7SE, and after a quick compsrison between the USB registers of
> the two chips, I noticed no major differences. The only thing was,
> that the SAM7SE has more endpoints and the pullup is integrated, but
> all the register and bit definitions are the same (except the bit
> controlling the pullup and the registers of the extra endpoints).
>
> Gábor
>
>   
I haven't tried it with anything other then the 7S, but if you're going 
to try it with the 7SE, get the latest CSV sources. There have been a 
few bug fixes to the driver in the last month or so. Let me know how it 
turns out - or if you have any problems. I've been looking at that 
driver recently.

- Frank Pagliughi

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

Re: [ECOS] AT91 USB driver

[Gábor Török]

I'm working with the CVS sources already, but first I have to type
some definitions for the SAM7SE, because the support for it isn't
included in the CVS version :(
Fortunately it is very similar to the SAM7S and the SAM7X :)

First I will only modify the pullup part of the driver, because I do
not need the extra endpoints. I can give you an SVN access, if you
would like to contribute (either to the USB specific or the SAM7SE
specific part).

Gábor Török

On Sat, Oct 4, 2008 at 10:39 PM, Frank Pagliughi wrote:
> Gábor Török wrote:
>>
>> Hi!
>>
>> Have anybody tried the AT91 USB driver with other controllers than the
>> AT91SAM7S (I only tried with this)? I would like to use it for the
>> AT91SAM7SE, and after a quick compsrison between the USB registers of
>> the two chips, I noticed no major differences. The only thing was,
>> that the SAM7SE has more endpoints and the pullup is integrated, but
>> all the register and bit definitions are the same (except the bit
>> controlling the pullup and the registers of the extra endpoints).
>>
>> Gábor
>>
>>
>
> I haven't tried it with anything other then the 7S, but if you're going to
> try it with the 7SE, get the latest CSV sources. There have been a few bug
> fixes to the driver in the last month or so. Let me know how it turns out -
> or if you have any problems. I've been looking at that driver recently.
>
> - Frank Pagliughi
>

--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

Re: [ECOS] AT91 USB driver

[Gábor Török]

Hi!

I'm testing the driver, but I did some modifications, because Atmel
datasheet says that before any other registers are written, the pullup
must be enabled. But the only consequence is some extra USB resets, so
I think that shouldn't rendered the USB unusable.
I included the extra endpoints in the USB driver, so they get disabled
correctly.

One stange thing was, that the Atmel datasheet said fo AT91SAM7SE,
that endpoint 0 (main control endpoint) FIFO is 64 bytes long, but in
any other devices it is 8 bytes long. I have written 64 bytes in the
USB driver for FIFO size. I was struggling to get driver working for
some days. The good thing was, that I found other bugs and partly
corrected that pullup thing at initialization. Today I had the idea to
reduce the FIFO size for EP0 to 8 bytes, and it worked. I will have a
look into the USB specification and the code to see if it is an error
in the Atmel datasheet, or is it a bug in the A91 USB driver or is it
a problem with my computer.

It works with SAM7SE, the original worked with SAM7 (this one should
also), and it should work with SAM7X and SAM7XC.

Until then, I commit it to my SVN and later today I will create a
public read access to allow testing. When it is well tested, I will
send the patches.

Gábor Török


On Sat, Oct 4, 2008 at 10:39 PM, Frank Pagliughi wrote:
> Gábor Török wrote:
>>
>> Hi!
>>
>> Have anybody tried the AT91 USB driver with other controllers than the
>> AT91SAM7S (I only tried with this)? I would like to use it for the
>> AT91SAM7SE, and after a quick compsrison between the USB registers of
>> the two chips, I noticed no major differences. The only thing was,
>> that the SAM7SE has more endpoints and the pullup is integrated, but
>> all the register and bit definitions are the same (except the bit
>> controlling the pullup and the registers of the extra endpoints).
>>
>> Gábor
>>
>>
>
> I haven't tried it with anything other then the 7S, but if you're going to
> try it with the 7SE, get the latest CSV sources. There have been a few bug
> fixes to the driver in the last month or so. Let me know how it turns out -
> or if you have any problems. I've been looking at that driver recently.
>
> - Frank Pagliughi
>

--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

Re: [ECOS] AT91 USB driver

[Frank Pagliughi]

Hello Gábor,

I looked through the Atmel summary and datasheets, and every place it's 
listed says that EP0 has a max size of 64 bytes. What value are you 
using for the "Max Packet Size" in the device descriptor?  If the driver 
is expecting 64, but the device descriptor tells the host to send 8 at a 
time, some packets will fly back and forth but it won't enumerate.  I've 
made that mistake before with other USB chips.

Frank


Gábor Török wrote:
> Hi!
>
> I'm testing the driver, but I did some modifications, because Atmel
> datasheet says that before any other registers are written, the pullup
> must be enabled. But the only consequence is some extra USB resets, so
> I think that shouldn't rendered the USB unusable.
> I included the extra endpoints in the USB driver, so they get disabled
> correctly.
>
> One stange thing was, that the Atmel datasheet said fo AT91SAM7SE,
> that endpoint 0 (main control endpoint) FIFO is 64 bytes long, but in
> any other devices it is 8 bytes long. I have written 64 bytes in the
> USB driver for FIFO size. I was struggling to get driver working for
> some days. The good thing was, that I found other bugs and partly
> corrected that pullup thing at initialization. Today I had the idea to
> reduce the FIFO size for EP0 to 8 bytes, and it worked. I will have a
> look into the USB specification and the code to see if it is an error
> in the Atmel datasheet, or is it a bug in the A91 USB driver or is it
> a problem with my computer.
>
> It works with SAM7SE, the original worked with SAM7 (this one should
> also), and it should work with SAM7X and SAM7XC.
>
> Until then, I commit it to my SVN and later today I will create a
> public read access to allow testing. When it is well tested, I will
> send the patches.
>
> Gábor Török
>
>
> On Sat, Oct 4, 2008 at 10:39 PM, Frank Pagliughi wrote:
>   
>> Gábor Török wrote:
>>     
>>> Hi!
>>>
>>> Have anybody tried the AT91 USB driver with other controllers than the
>>> AT91SAM7S (I only tried with this)? I would like to use it for the
>>> AT91SAM7SE, and after a quick compsrison between the USB registers of
>>> the two chips, I noticed no major differences. The only thing was,
>>> that the SAM7SE has more endpoints and the pullup is integrated, but
>>> all the register and bit definitions are the same (except the bit
>>> controlling the pullup and the registers of the extra endpoints).
>>>
>>> Gábor
>>>
>>>
>>>       
>> I haven't tried it with anything other then the 7S, but if you're going to
>> try it with the 7SE, get the latest CSV sources. There have been a few bug
>> fixes to the driver in the last month or so. Let me know how it turns out -
>> or if you have any problems. I've been looking at that driver recently.
>>
>> - Frank Pagliughi
>>
>>     
>
>   

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

Re: [ECOS] AT91 USB driver

[Gábor Török]

Hello Frank,

Thank you for the suggestion, I noticed that for the USB serial slave,
I must set this parameter manually, so I tried setting 64 bytes both
in the AT91 driver and in the USB serial, but it didn't help. It was
better than before, because it started the enumeration, and the device
descriptor was received by the PC, as my USB debug tool said (earlier
I saw just zeroes), but that was all, and it didn't finish the
enumeration. Until I'll have time to figure out the problem, I'll
leave it at 8 bytes. Everybody is welcome to solve it.

Read-only access:
svn://tgabor.no-ip.com/eCos/trunk/packages
user, pass: ecoslist, ecoslist

To get a read-write access, write to me (not to the eCos mailing list).

Gábor Török

On Wed, Oct 22, 2008 at 2:42 AM, Frank Pagliughi
<fpagliughi@mindspring.com> wrote:
> Hello Gábor,
>
> I looked through the Atmel summary and datasheets, and every place it's
> listed says that EP0 has a max size of 64 bytes. What value are you using
> for the "Max Packet Size" in the device descriptor?  If the driver is
> expecting 64, but the device descriptor tells the host to send 8 at a time,
> some packets will fly back and forth but it won't enumerate.  I've made that
> mistake before with other USB chips.
>
> Frank
>
>
> Gábor Török wrote:
>>
>> Hi!
>>
>> I'm testing the driver, but I did some modifications, because Atmel
>> datasheet says that before any other registers are written, the pullup
>> must be enabled. But the only consequence is some extra USB resets, so
>> I think that shouldn't rendered the USB unusable.
>> I included the extra endpoints in the USB driver, so they get disabled
>> correctly.
>>
>> One stange thing was, that the Atmel datasheet said fo AT91SAM7SE,
>> that endpoint 0 (main control endpoint) FIFO is 64 bytes long, but in
>> any other devices it is 8 bytes long. I have written 64 bytes in the
>> USB driver for FIFO size. I was struggling to get driver working for
>> some days. The good thing was, that I found other bugs and partly
>> corrected that pullup thing at initialization. Today I had the idea to
>> reduce the FIFO size for EP0 to 8 bytes, and it worked. I will have a
>> look into the USB specification and the code to see if it is an error
>> in the Atmel datasheet, or is it a bug in the A91 USB driver or is it
>> a problem with my computer.
>>
>> It works with SAM7SE, the original worked with SAM7 (this one should
>> also), and it should work with SAM7X and SAM7XC.
>>
>> Until then, I commit it to my SVN and later today I will create a
>> public read access to allow testing. When it is well tested, I will
>> send the patches.
>>
>> Gábor Török
>>
>>
>> On Sat, Oct 4, 2008 at 10:39 PM, Frank Pagliughi wrote:
>>
>>>
>>> Gábor Török wrote:
>>>
>>>>
>>>> Hi!
>>>>
>>>> Have anybody tried the AT91 USB driver with other controllers than the
>>>> AT91SAM7S (I only tried with this)? I would like to use it for the
>>>> AT91SAM7SE, and after a quick compsrison between the USB registers of
>>>> the two chips, I noticed no major differences. The only thing was,
>>>> that the SAM7SE has more endpoints and the pullup is integrated, but
>>>> all the register and bit definitions are the same (except the bit
>>>> controlling the pullup and the registers of the extra endpoints).
>>>>
>>>> Gábor
>>>>
>>>>
>>>>
>>>
>>> I haven't tried it with anything other then the 7S, but if you're going
>>> to
>>> try it with the 7SE, get the latest CSV sources. There have been a few
>>> bug
>>> fixes to the driver in the last month or so. Let me know how it turns out
>>> -
>>> or if you have any problems. I've been looking at that driver recently.
>>>
>>> - Frank Pagliughi
>>>
>>>
>>
>>
>

--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

[ECOS] AT91 USB driver

[Gábor Török]

Hi!

I'm I seeing right, that the driver ignores USB VBUS state and it
assumes that it is always present (as in the case of a bus powered
device)?

Gábor

--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss