There is a bug in Samsung's UDC driver, which is completely disabling
the USB device when a custom UDC command is used.
Following patch seems to get the right?behavior (e.g. enabling pull-up
instead of disabling then Vcc is applied). Also I'm not in mailing
list so please CC me.
Greetingz,
Vilo
Signed-off-by: Viliam Mateicka <[email protected]>
--- drivers/usb/gadget/s3c2410_udc.c.orig ? ? ? 2011-05-19
06:06:34.000000000 +0200
+++ drivers/usb/gadget/s3c2410_udc.c ? ?2011-06-16 08:57:19.000000000 +0200
@@ -1566,7 +1566,7 @@ static void s3c2410_udc_command(enum s3c
? ? ? ? ? ? ? ? return;
? ? ? ? if (udc_info->udc_command) {
- ? ? ? ? ? ? ? udc_info->udc_command(S3C2410_UDC_P_DISABLE);
+ ? ? ? ? ? ? ? udc_info->udc_command(cmd);
? ? ? ? } else if (gpio_is_valid(udc_info->pullup_pin)) {
? ? ? ? ? ? ? ? int value;
So, will somebody at least react/reply (or just fix) ?
Vilo
On Thu, Jun 16, 2011 at 10:04 AM, Viliam Mateicka
<[email protected]> wrote:
> There is a bug in Samsung's UDC driver, which is completely disabling
> the USB device when a custom UDC command is used.
> Following patch seems to get the right?behavior (e.g. enabling pull-up
> instead of disabling then Vcc is applied). Also I'm not in mailing
> list so please CC me.
>
> Greetingz,
> Vilo
>
> Signed-off-by: Viliam Mateicka <[email protected]>
>
> --- drivers/usb/gadget/s3c2410_udc.c.orig ? ? ? 2011-05-19
> 06:06:34.000000000 +0200
> +++ drivers/usb/gadget/s3c2410_udc.c ? ?2011-06-16 08:57:19.000000000 +0200
> @@ -1566,7 +1566,7 @@ static void s3c2410_udc_command(enum s3c
> ? ? ? ? ? ? ? ? return;
> ? ? ? ? if (udc_info->udc_command) {
> - ? ? ? ? ? ? ? udc_info->udc_command(S3C2410_UDC_P_DISABLE);
> + ? ? ? ? ? ? ? udc_info->udc_command(cmd);
> ? ? ? ? } else if (gpio_is_valid(udc_info->pullup_pin)) {
> ? ? ? ? ? ? ? ? int value;
>
On Mon, 20 Jun 2011, Viliam Mateicka wrote:
> So, will somebody at least react/reply (or just fix) ?
Have some patience. You did not address your email message to either
of the driver's authors or to the current Gadget maintainer. Also,
David Brownell is deceased and Greg KH is on vacation.
Alan Stern
Hi,
On Mon, Jun 20, 2011 at 11:07:51AM -0400, Alan Stern wrote:
> On Mon, 20 Jun 2011, Viliam Mateicka wrote:
>
> > So, will somebody at least react/reply (or just fix) ?
>
> Have some patience. You did not address your email message to either
> of the driver's authors or to the current Gadget maintainer. Also,
> David Brownell is deceased and Greg KH is on vacation.
besides, the patch formatting is wrong. That was noted before and the
reply was less than acceptably polite, so unless the patch comes
Perfect(TM), I'm not looking into it.
--
balbi
I've been exactly following
http://www.kernel.org/doc/Documentation/SubmittingPatches:
To create a patch for a single file, it is often sufficient to do:
SRCTREE= linux-2.6
MYFILE= drivers/net/mydriver.c
cd $SRCTREE
cp $MYFILE $MYFILE.orig
vi $MYFILE # make your change
cd ..
diff -up $SRCTREE/$MYFILE{.orig,} > /tmp/patch
And that's what I've done. What do you think is wrong with it ???
Vilo
On Mon, Jun 20, 2011 at 5:15 PM, Felipe Balbi <[email protected]> wrote:
> Hi,
>
> On Mon, Jun 20, 2011 at 11:07:51AM -0400, Alan Stern wrote:
>> On Mon, 20 Jun 2011, Viliam Mateicka wrote:
>>
>> > So, will somebody at least react/reply (or just fix) ?
>>
>> Have some patience. ?You did not address your email message to either
>> of the driver's authors or to the current Gadget maintainer. ?Also,
>> David Brownell is deceased and Greg KH is on vacation.
>
> besides, the patch formatting is wrong. That was noted before and the
> reply was less than acceptably polite, so unless the patch comes
> Perfect(TM), I'm not looking into it.
>
> --
> balbi
>
Hi,
On Mon, Jun 20, 2011 at 05:20:02PM +0200, Skacore Systems wrote:
> I've been exactly following
> http://www.kernel.org/doc/Documentation/SubmittingPatches:
>
> To create a patch for a single file, it is often sufficient to do:
>
> SRCTREE= linux-2.6
> MYFILE= drivers/net/mydriver.c
>
> cd $SRCTREE
> cp $MYFILE $MYFILE.orig
> vi $MYFILE # make your change
> cd ..
> diff -up $SRCTREE/$MYFILE{.orig,} > /tmp/patch
>
> And that's what I've done. What do you think is wrong with it ???
patch should be in 'p1' format to start with, not 'p0'!!
Your commit log shouldn't include your "regards" or anything like that,
or that will go into the git history.
> Subject: [PATCH] s3c2410_udc: fix custom UDC command handling
this should be something like:
[PATCH] usb: gadget: s3c2410_udc: fix custom UDC command handling
we generally put the subsystem on the subject to help maintainers
filter patches related to them.
> There is a bug in Samsung's UDC driver, which is completely disabling
> the USB device when a custom UDC command is used.
> Following patch seems to get the right?behavior (e.g. enabling pull-up
> instead of disabling then Vcc is applied).
"seems to get the right behavior" makes me wonder if you really tested
this properly. Either it fixes the problem, or it doesn't. Have you
tried with other possible commands to see if also helps ?
> Also I'm not in mailing list so please CC me.
this would go to commit log, shouldn't be here.
> Greetingz,
> Vilo
this shouldn't be here either.
> Signed-off-by: Viliam Mateicka <[email protected]>
there should be a tear line here. (---)
> --- drivers/usb/gadget/s3c2410_udc.c.orig ? ? ? 2011-05-19
> 06:06:34.000000000 +0200
this line break is bogus.
you could have looked at any mailing list archive for properly formatted
patches, but instead you chose to be unpolite with the persons who are
giving you a "service" free of charge.
At [1] you can find an example of a proper patch
[1] http://marc.info/?l=linux-usb&m=130833192703040&w=2
--
balbi
Ok I'm giving it up. This is just a non-sence.
I've followed the exact rules written there, so tell me how the patch
could be wrong.
The rules are obsolete or you made up a new ones ?
Regarding to the testing. If you've at least looked in to the patch
(and you probably not giving a stupid questions about it!)
you would know that this is 100% bug and you don't even have to
compile to see that fix is actually working.
And of course I've tested it on our production kernel.
Look somehow it's gonna fix it. It must not be you,
someone will just (I hope Samsung's people do care about their stuff) do it.
The amount of time I and you've put into this discussion is just pure waste.
Vilo
On Mon, Jun 20, 2011 at 5:36 PM, Felipe Balbi <[email protected]> wrote:
> Hi,
>
> On Mon, Jun 20, 2011 at 05:20:02PM +0200, Skacore Systems wrote:
>> I've been exactly following
>> http://www.kernel.org/doc/Documentation/SubmittingPatches:
>>
>> To create a patch for a single file, it is often sufficient to do:
>>
>> ? ? ? SRCTREE= linux-2.6
>> ? ? ? MYFILE= ?drivers/net/mydriver.c
>>
>> ? ? ? cd $SRCTREE
>> ? ? ? cp $MYFILE $MYFILE.orig
>> ? ? ? vi $MYFILE ? ? ?# make your change
>> ? ? ? cd ..
>> ? ? ? diff -up $SRCTREE/$MYFILE{.orig,} > /tmp/patch
>>
>> And that's what I've done. What do you think is wrong with it ???
>
> patch should be in 'p1' format to start with, not 'p0'!!
>
> Your commit log shouldn't include your "regards" or anything like that,
> or that will go into the git history.
>
>> Subject: [PATCH] s3c2410_udc: fix custom UDC command handling
>
> this should be something like:
>
> [PATCH] usb: gadget: s3c2410_udc: fix custom UDC command handling
>
> we generally put the subsystem on the subject to help maintainers
> filter patches related to them.
>
>> There is a bug in Samsung's UDC driver, which is completely disabling
>> the USB device when a custom UDC command is used.
>> Following patch seems to get the right?behavior (e.g. enabling pull-up
>> instead of disabling then Vcc is applied).
>
> "seems to get the right behavior" makes me wonder if you really tested
> this properly. Either it fixes the problem, or it doesn't. Have you
> tried with other possible commands to see if also helps ?
>
>> Also I'm not in mailing list so please CC me.
>
> this would go to commit log, shouldn't be here.
>
>> Greetingz,
>> Vilo
>
> this shouldn't be here either.
>
>> Signed-off-by: Viliam Mateicka <[email protected]>
>
> there should be a tear line here. (---)
>
>> --- drivers/usb/gadget/s3c2410_udc.c.orig ? ? ? 2011-05-19
>> 06:06:34.000000000 +0200
>
> this line break is bogus.
>
> you could have looked at any mailing list archive for properly formatted
> patches, but instead you chose to be unpolite with the persons who are
> giving you a "service" free of charge.
>
> At [1] you can find an example of a proper patch
>
> [1] http://marc.info/?l=linux-usb&m=130833192703040&w=2
>
> --
> balbi
>
Oh and I've finally cloned the mainline git. So here it's the last try.
I hope the TI people will not sabotage the Samsung's drivers any way longer !!!
Vilo
diff --git a/drivers/usb/gadget/s3c2410_udc.c b/drivers/usb/gadget/s3c2410_udc.c
index 100f263..b6e1ca5 100644
--- a/drivers/usb/gadget/s3c2410_udc.c
+++ b/drivers/usb/gadget/s3c2410_udc.c
@@ -1567,7 +1567,7 @@ static void s3c2410_udc_command(enum
s3c2410_udc_cmd_e cmd)
return;
if (udc_info->udc_command) {
- udc_info->udc_command(S3C2410_UDC_P_DISABLE);
+ udc_info->udc_command(cmd);
} else if (gpio_is_valid(udc_info->pullup_pin)) {
int value;
On Mon, Jun 20, 2011 at 5:48 PM, Skacore Systems
<[email protected]> wrote:
> Ok I'm giving it up. This is just a non-sence.
> I've followed the exact rules written there, so tell me how the patch
> could be wrong.
> The rules are obsolete or you made up a new ones ?
> Regarding to the testing. If you've at least looked in to the patch
> (and you probably not giving a stupid questions about it!)
> you would know that this is 100% bug and you don't even have to
> compile to see that fix is actually working.
> And of course I've tested it on our production kernel.
> Look somehow it's gonna fix it. It must not be you,
> someone will just (I hope Samsung's people do care about their stuff) do it.
> The amount of time I and you've put into this discussion is just pure waste.
>
> Vilo
>
>
> On Mon, Jun 20, 2011 at 5:36 PM, Felipe Balbi <[email protected]> wrote:
>> Hi,
>>
>> On Mon, Jun 20, 2011 at 05:20:02PM +0200, Skacore Systems wrote:
>>> I've been exactly following
>>> http://www.kernel.org/doc/Documentation/SubmittingPatches:
>>>
>>> To create a patch for a single file, it is often sufficient to do:
>>>
>>> ? ? ? SRCTREE= linux-2.6
>>> ? ? ? MYFILE= ?drivers/net/mydriver.c
>>>
>>> ? ? ? cd $SRCTREE
>>> ? ? ? cp $MYFILE $MYFILE.orig
>>> ? ? ? vi $MYFILE ? ? ?# make your change
>>> ? ? ? cd ..
>>> ? ? ? diff -up $SRCTREE/$MYFILE{.orig,} > /tmp/patch
>>>
>>> And that's what I've done. What do you think is wrong with it ???
>>
>> patch should be in 'p1' format to start with, not 'p0'!!
>>
>> Your commit log shouldn't include your "regards" or anything like that,
>> or that will go into the git history.
>>
>>> Subject: [PATCH] s3c2410_udc: fix custom UDC command handling
>>
>> this should be something like:
>>
>> [PATCH] usb: gadget: s3c2410_udc: fix custom UDC command handling
>>
>> we generally put the subsystem on the subject to help maintainers
>> filter patches related to them.
>>
>>> There is a bug in Samsung's UDC driver, which is completely disabling
>>> the USB device when a custom UDC command is used.
>>> Following patch seems to get the right?behavior (e.g. enabling pull-up
>>> instead of disabling then Vcc is applied).
>>
>> "seems to get the right behavior" makes me wonder if you really tested
>> this properly. Either it fixes the problem, or it doesn't. Have you
>> tried with other possible commands to see if also helps ?
>>
>>> Also I'm not in mailing list so please CC me.
>>
>> this would go to commit log, shouldn't be here.
>>
>>> Greetingz,
>>> Vilo
>>
>> this shouldn't be here either.
>>
>>> Signed-off-by: Viliam Mateicka <[email protected]>
>>
>> there should be a tear line here. (---)
>>
>>> --- drivers/usb/gadget/s3c2410_udc.c.orig ? ? ? 2011-05-19
>>> 06:06:34.000000000 +0200
>>
>> this line break is bogus.
>>
>> you could have looked at any mailing list archive for properly formatted
>> patches, but instead you chose to be unpolite with the persons who are
>> giving you a "service" free of charge.
>>
>> At [1] you can find an example of a proper patch
>>
>> [1] http://marc.info/?l=linux-usb&m=130833192703040&w=2
>>
>> --
>> balbi
>>
>
Hi,
On Mon, Jun 20, 2011 at 05:57:31PM +0200, Skacore Systems wrote:
> Oh and I've finally cloned the mainline git. So here it's the last
> try. I hope the TI people will not sabotage the Samsung's drivers any
> way longer !!!
Now you're just making an ass of yourself.
what do you think you will get making such kind of statements ? I just
happen to be hired by TI, there's no such thing as TI trying to sabotage
Samsung's drivers.
I'll fix this one out myself just to make you stop bothering me, but
don't expect me to put your patches in my priority list. If you really
think TI is sabotaging samsung's drivers, look at the big re-factor we
are doing on the entire Gadget framework, involving not only Samsung's
drivers but all the others.
Even though you cloned the tree you still make a mistake which will
prevent me from applying the patch without manually editting it.
On top of that, you decided not to follow our netiquette or our patch
formatting rules just because you think it's "nonsense". If you really
think that, why do you even bother sending the patch to this mailing
list ?
> diff --git a/drivers/usb/gadget/s3c2410_udc.c b/drivers/usb/gadget/s3c2410_udc.c
> index 100f263..b6e1ca5 100644
> --- a/drivers/usb/gadget/s3c2410_udc.c
> +++ b/drivers/usb/gadget/s3c2410_udc.c
> @@ -1567,7 +1567,7 @@ static void s3c2410_udc_command(enum
> s3c2410_udc_cmd_e cmd)
there shouldn't be this line break here.
> return;
>
> if (udc_info->udc_command) {
> - udc_info->udc_command(S3C2410_UDC_P_DISABLE);
> + udc_info->udc_command(cmd);
> } else if (gpio_is_valid(udc_info->pullup_pin)) {
> int value;
tabs where converted into spaces too.
Next time, instead of been such an ass, spend some time learning the
process we use to do development. You'll see things will go smoother.
--
balbi
On Mon, Jun 20, 2011 at 9:27 PM, Skacore Systems
<[email protected]> wrote:
> Oh and I've finally cloned the mainline git. So here it's the last try.
> I hope the TI people will not sabotage the Samsung's drivers any way longer !!!
>
> Vilo
If you already cloned the git, it'd do good to have a quick look at
the output of 'git log'
and see how patches are supposed to be. And hopefully you'll realize
the following
wouldn't look much useful if someone did 'git log' over your patch.
"
Oh and I've finally cloned the mainline git. So here it's the last try.
I hope the TI people will not sabotage the Samsung's drivers any way longer !!!
"
Yeah I think that one line break and tabs are sabotaging your work.
So stop being an 8====O - hole. No wonder if you are paid by TI that
one line patch into the upstream takes you so long to comprehend.
Vilo
On Mon, Jun 20, 2011 at 6:06 PM, Felipe Balbi <[email protected]> wrote:
> Hi,
>
> On Mon, Jun 20, 2011 at 05:57:31PM +0200, Skacore Systems wrote:
>> Oh and I've finally cloned the mainline git. So here it's the last
>> try. I hope the TI people will not sabotage the Samsung's drivers any
>> way longer !!!
>
> Now you're just making an ass of yourself.
>
> what do you think you will get making such kind of statements ? I just
> happen to be hired by TI, there's no such thing as TI trying to sabotage
> Samsung's drivers.
>
> I'll fix this one out myself just to make you stop bothering me, but
> don't expect me to put your patches in my priority list. If you really
> think TI is sabotaging samsung's drivers, look at the big re-factor we
> are doing on the entire Gadget framework, involving not only Samsung's
> drivers but all the others.
>
> Even though you cloned the tree you still make a mistake which will
> prevent me from applying the patch without manually editting it.
>
> On top of that, you decided not to follow our netiquette or our patch
> formatting rules just because you think it's "nonsense". If you really
> think that, why do you even bother sending the patch to this mailing
> list ?
>
>> diff --git a/drivers/usb/gadget/s3c2410_udc.c b/drivers/usb/gadget/s3c2410_udc.c
>> index 100f263..b6e1ca5 100644
>> --- a/drivers/usb/gadget/s3c2410_udc.c
>> +++ b/drivers/usb/gadget/s3c2410_udc.c
>> @@ -1567,7 +1567,7 @@ static void s3c2410_udc_command(enum
>> s3c2410_udc_cmd_e cmd)
>
> there shouldn't be this line break here.
>
>> ? ? ? ? ? ? ? ? return;
>>
>> ? ? ? ? if (udc_info->udc_command) {
>> - ? ? ? ? ? ? ? udc_info->udc_command(S3C2410_UDC_P_DISABLE);
>> + ? ? ? ? ? ? ? udc_info->udc_command(cmd);
>> ? ? ? ? } else if (gpio_is_valid(udc_info->pullup_pin)) {
>> ? ? ? ? ? ? ? ? int value;
>
> tabs where converted into spaces too.
>
> Next time, instead of been such an ass, spend some time learning the
> process we use to do development. You'll see things will go smoother.
>
> --
> balbi
>
Hi,
On Mon, Jun 20, 2011 at 06:41:59PM +0200, Skacore Systems wrote:
> Yeah I think that one line break and tabs are sabotaging your work.
> So stop being an 8====O - hole. No wonder if you are paid by TI that
> one line patch into the upstream takes you so long to comprehend.
you are this close for me to drop your fix and add you to the shitlist.
--
balbi
I wonder then, who applies the patch first. One ass can't stop
progress (was there some progress afterall ?) because of one
line-break (which git produces by default afterall) and several
non-conforming whitespace conversions. I think you should stop
replying to me, because I'm kinda starting to like trolling afterall.
Vilo
On Mon, Jun 20, 2011 at 7:05 PM, Felipe Balbi <[email protected]> wrote:
> Hi,
>
> On Mon, Jun 20, 2011 at 06:41:59PM +0200, Skacore Systems wrote:
>> Yeah I think that one line break and tabs are sabotaging your work.
>> So stop being an 8====O - hole. No wonder if you are paid by TI that
>> one line patch into the upstream takes you so long to comprehend.
>
> you are this close for me to drop your fix and add you to the shitlist.
>
> --
> balbi
>
Hi,
On Mon, Jun 20, 2011 at 07:27:18PM +0200, Viliam Mateicka wrote:
> I wonder then, who applies the patch first. One ass can't stop
> progress (was there some progress afterall ?) because of one
> line-break (which git produces by default afterall) and several
> non-conforming whitespace conversions. I think you should stop
> replying to me, because I'm kinda starting to like trolling afterall.
Look, maybe you're just new to git/linux and that's fine. The only
problem we had from day one is that you have been a complete dick when
replying.
If you go that road, that's what you'll get in return.
Just so you see the properly formatted patch, I'm quoting below. See
that there's no line breaks introduced by git.
commit c27c9b5312f4b2a7e66ee772943355d77fc2c550
Author: Viliam Mateicka <[email protected]>
Date: Mon Jun 20 19:08:27 2011 +0300
usb: gadget: s3c2410_udc: fix custom UDC command handling
There is a bug in Samsung's UDC driver, which is
completely disabling the USB device when a custom
UDC command is used.
Following patch seems to get the right?behavior
(e.g. enabling pull-up instead of disabling then
Vcc is applied).
Signed-off-by: Viliam Mateicka <[email protected]>
[ [email protected] : manually re-created the patch because
author was unable to send a proper one ]
Signed-off-by: Felipe Balbi <[email protected]>
diff --git a/drivers/usb/gadget/s3c2410_udc.c b/drivers/usb/gadget/s3c2410_udc.c
index 100f263..b6e1ca5 100644
--- a/drivers/usb/gadget/s3c2410_udc.c
+++ b/drivers/usb/gadget/s3c2410_udc.c
@@ -1567,7 +1567,7 @@ static void s3c2410_udc_command(enum s3c2410_udc_cmd_e cmd)
return;
if (udc_info->udc_command) {
- udc_info->udc_command(S3C2410_UDC_P_DISABLE);
+ udc_info->udc_command(cmd);
} else if (gpio_is_valid(udc_info->pullup_pin)) {
int value;
(if you see any line breaks that's your mailer, it's better to check
from the archives [1])
[1] http://marc.info/?l=linux-usb
--
balbi
Look. I didn't wanna spend more time with the issue because for me -
it was solved. It was just the stupid byrocratic system you've
enforced to minimize your work - which I would say its complete
non-sence if it could ever have a problem with whitespace or line
break. I understand that on the source level every character it's
important to the look of the code, but at least when you try to force
somebody to use some rules (which most of it at my copy were applied
fine just fine by myself), at least use a tool which take a care about
it (the look of the kernel code). Just look how much time it took to
upstream a little change (valid on the first look), just because of my
ignorance to some the formating rules (linebreaks and tabs namely - if
you have a proper editor which takes care about it.) and a faulty
mailer (I).
Anyway, if the Documentation/SubmittingPatches would mention git
format-patch and git send-email I would have use it and there would'nt
be any trouble using faulty mailer. Instead as a first option there
is solution which almost everytime produces such a problems.
So many thanks, for doing your job and keeping the bad asses using
wrong tools out of the formating heaven. I could imagine that this
would be hard (to get one line into the kernel) but still haven't
thought that hard. I would like to thank you for your patience and the
exact amount of right doingness, while showing me what's wrong instead
of telling me how to do it properly.
Also liked the diff between our patches:
player tmp # diff -ur wastetime wastetime2
--- wastetime 2011-06-20 20:33:25.000000000 +0200
+++ wastetime2 2011-06-20 20:30:21.000000000 +0200
@@ -2,11 +2,12 @@
index 100f263..b6e1ca5 100644
--- a/drivers/usb/gadget/s3c2410_udc.c
+++ b/drivers/usb/gadget/s3c2410_udc.c
-@@ -1567,7 +1567,7 @@ static void s3c2410_udc_command(enum
s3c2410_udc_cmd_e cmd)
- return;
-
- if (udc_info->udc_command) {
-- udc_info->udc_command(S3C2410_UDC_P_DISABLE);
-+ udc_info->udc_command(cmd);
- } else if (gpio_is_valid(udc_info->pullup_pin)) {
- int value;
+@@ -1567,7 +1567,7 @@ static void s3c2410_udc_command(enum
+s3c2410_udc_cmd_e cmd)
+ return;
+
+ if (udc_info->udc_command) {
+- udc_info->udc_command(S3C2410_UDC_P_DISABLE);
++ udc_info->udc_command(cmd);
+ } else if (gpio_is_valid(udc_info->pullup_pin)) {
+ int value;
Vilo
On Mon, Jun 20, 2011 at 7:34 PM, Felipe Balbi <[email protected]> wrote:
> Hi,
>
> On Mon, Jun 20, 2011 at 07:27:18PM +0200, Viliam Mateicka wrote:
>> I wonder then, who applies the patch first. One ass can't stop
>> progress (was there some progress afterall ?) because of one
>> line-break (which git produces by default afterall) and several
>> non-conforming whitespace conversions. I think you should stop
>> replying to me, because I'm kinda starting to like trolling afterall.
>
> Look, maybe you're just new to git/linux and that's fine. The only
> problem we had from day one is that you have been a complete dick when
> replying.
>
> If you go that road, that's what you'll get in return.
>
> Just so you see the properly formatted patch, I'm quoting below. See
> that there's no line breaks introduced by git.
>
> commit c27c9b5312f4b2a7e66ee772943355d77fc2c550
> Author: Viliam Mateicka <[email protected]>
> Date: ? Mon Jun 20 19:08:27 2011 +0300
>
> ? ?usb: gadget: s3c2410_udc: fix custom UDC command handling
>
> ? ?There is a bug in Samsung's UDC driver, which is
> ? ?completely disabling the USB device when a custom
> ? ?UDC command is used.
>
> ? ?Following patch seems to get the right?behavior
> ? ?(e.g. enabling pull-up instead of disabling then
> ? ?Vcc is applied).
>
> ? ?Signed-off-by: Viliam Mateicka <[email protected]>
>
> ? ?[ [email protected] : manually re-created the patch because
> ? ? ? ?author was unable to send a proper one ]
>
> ? ?Signed-off-by: Felipe Balbi <[email protected]>
>
> diff --git a/drivers/usb/gadget/s3c2410_udc.c b/drivers/usb/gadget/s3c2410_udc.c
> index 100f263..b6e1ca5 100644
> --- a/drivers/usb/gadget/s3c2410_udc.c
> +++ b/drivers/usb/gadget/s3c2410_udc.c
> @@ -1567,7 +1567,7 @@ static void s3c2410_udc_command(enum s3c2410_udc_cmd_e cmd)
> ? ? ? ? ? ? ? ?return;
>
> ? ? ? ?if (udc_info->udc_command) {
> - ? ? ? ? ? ? ? udc_info->udc_command(S3C2410_UDC_P_DISABLE);
> + ? ? ? ? ? ? ? udc_info->udc_command(cmd);
> ? ? ? ?} else if (gpio_is_valid(udc_info->pullup_pin)) {
> ? ? ? ? ? ? ? ?int value;
>
>
> (if you see any line breaks that's your mailer, it's better to check
> from the archives [1])
>
> [1] http://marc.info/?l=linux-usb
>
> --
> balbi
>