Re: [PATCH] musb: fix DaVinci build again

Re: [PATCH] musb: fix DaVinci build again

[Sergei Shtylyov]

Sergei Shtylyov wrote:
> Hello, I wrote:
>
>>>> The last Kevin Hilman's patch seems to have been broken on its way 
>>>> to Greg's
>>>> tree and also probably clashed with yet another patch. Fix DaVinci 
>>>> build again.
>>>> Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
>>>> ---
>>>> Felipe, I have to ask you again to be more careful with patches...
>>
>>>    11 days have passed since this has benn sent out and I'm still 
>>> not seeing this 2.6.28 *hot* fix in the MUSB git tree. Greg, could 
>>> you queue this urgently, bypassing the MUSB tree?
>>
>>   I'll be happy if it gets combined with the following patch in your 
>> tree:
>>
>> http://www.kernel.org/pub/linux/kernel/people/gregkh/gregkh-2.6/patches/usb/usb-musb-build-fixes-for-davinci.patch 
>>
>>
>> fallout of which I'm fixing. Though it would've been better yet if 
>> you just dropped that totally broken patch. I could've send a 
>> replacment then...
>
>   The patch I liked to above fixes nothing and only adds more 
> breakage. Compare it with the original Kevin's patch:
>
> http://marc.info/?l=linux-usb&m=122608723815504

   I'm seeing Felipe returning to the list but still ignoring this 
thread so far.
   For how much longer we'll have to wait for this issue to be resolved? 
2.6.28 is 2 days away (at most).
   Greg, what's the problem with dropping obviously broken patch?

WBR, Sergei

Re: [PATCH] musb: fix DaVinci build again

[David Brownell]

On Tuesday 23 December 2008, Sergei Shtylyov wrote:
>    For how much longer we'll have to wait for this issue to be resolved? 
> 2.6.28 is 2 days away (at most).

Yeah, I sent a DaVinci build fix patch in plenty of time to
be merged for *2.6.27* (!!!!) but still the key parts are
missing, and have not managed to make it into mainline.

Re: [PATCH] musb: fix DaVinci build again

[Sergei Shtylyov]

Hello.

David Brownell wrote:

>>    For how much longer we'll have to wait for this issue to be resolved? 
>> 2.6.28 is 2 days away (at most).
>>     
>
> Yeah, I sent a DaVinci build fix patch in plenty of time to
> be merged for *2.6.27* (!!!!) but still the key parts are
> missing, and have not managed to make it into mainline.

  Try breakinjg the patches into smaller pieces next time, so they may 
get thru the maintainer's sieve more easily. :-)

WBR, Sergei

Re: [PATCH] musb: fix DaVinci build again

[David Brownell]

On Tuesday 23 December 2008, Sergei Shtylyov wrote:
> > Yeah, I sent a DaVinci build fix patch in plenty of time to
> > be merged for *2.6.27* (!!!!) but still the key parts are
> > missing, and have not managed to make it into mainline.
> 
>   Try breakinjg the patches into smaller pieces next time, so they may 
> get thru the maintainer's sieve more easily. :-)

That wasn't the issue back then.  If you look at

  c767c1c6f1febbd1351cc152bba6e37889322d17

which got merged after 2.6.27 shipped, you'll notice that
someone stripped out the following parts from the original
patch, which were even referenced in the patch comment.

This worked as submitted, but then got broken during merge;
and fixes have been re-sent several times already.

Plus, some other stuff has merged which prevents linking.
Fixes for those parts have been re-sent several times too.

- Dave

============= CUT HERE
From: David Brownell <dbrownell@users.sourceforge.net>

Build fixes that somehow got stripped out of the earlier build
fix patch ...

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
---
 drivers/usb/musb/davinci.c |    7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

--- a/drivers/usb/musb/davinci.c
+++ b/drivers/usb/musb/davinci.c
@@ -31,10 +31,11 @@
 #include <linux/clk.h>
 #include <linux/io.h>
 #include <linux/gpio.h>
+#include <linux/gpio.h>
+
+#include <mach/hardware.h>
+#include <mach/memory.h>
 
-#include <asm/arch/hardware.h>
-#include <asm/arch/memory.h>
-#include <asm/arch/gpio.h>
 #include <asm/mach-types.h>
 
 #include "musb_core.h"

Re: [PATCH] musb: fix DaVinci build again

[Felipe Balbi]

Hi,

just a note before a comment. I'm gonna be on vacations during this and
next week. Trying to get better from a really terrible flu so I won't be
hacking too much during this time.

Anyways, once comment below.

On Tue, Dec 23, 2008 at 05:07:47PM -0800, David Brownell wrote:
> From: David Brownell <dbrownell@users.sourceforge.net>
> 
> Build fixes that somehow got stripped out of the earlier build
> fix patch ...
> 
> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
> ---
>  drivers/usb/musb/davinci.c |    7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> --- a/drivers/usb/musb/davinci.c
> +++ b/drivers/usb/musb/davinci.c
> @@ -31,10 +31,11 @@
>  #include <linux/clk.h>
>  #include <linux/io.h>
>  #include <linux/gpio.h>
> +#include <linux/gpio.h>

you're including <linux/gpio.h> twice.

-- 
balbi

Re: [PATCH] musb: fix DaVinci build again

[David Brownell]

On Tuesday 23 December 2008, Felipe Balbi wrote:
> >  #include <linux/gpio.h>
> > +#include <linux/gpio.h>
> 
> you're including <linux/gpio.h> twice.

Yeah, I suppose I just edited the patch to
delete segments that *did* get merged.  It
looks like that change to <linux/gpio.h> got
included, and "patch" didn't complain ...
leaving the puzzle of just why the *rest* of
that pre-2.6.27 build fix still hasn't merged.

A double inclusion is a minor-but-harmless
annoyance.  Those <asm/arch/...> ==> <mach/...>
header changes were mostly merged by 2.6.27-rc2,
but ones for this particular file got stripped
out.  More than a small bit puzzling; I can
understand patches getting dropped or lost,
but not getting mangled like that.

When I verify that I have a kernel booting
on the DM6446 EVM in mainline, I'll resend
both of the patches.  They are IMO overdue
to merge.

Re: [PATCH] musb: fix DaVinci build again

[David Brownell]

Greg, I suggest you just use the appended patch to *REPLACE* the
patch now in your gregkh-04-usb queue:

   usb-musb-build-fixes-for-davinci.patch

Net effect of including this plus the other patches in
that queue is to achieve what the original 2.6.27-rc(3?)
build fix patch would have done, if it were merged right.

- Dave

===========
From: Kevin Hilman <khilman@deeprootsystems.com>

 - update includes after asm/arch --> mach headers move

[ dbrownell@users.sourceforge.net: just fix mismerge of my 2.6.27-rc patch ]

Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>
Signed-off-by: Felipe Balbi <felipe.balbi@nokia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

---
 drivers/usb/musb/davinci.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

--- a/drivers/usb/musb/davinci.c
+++ b/drivers/usb/musb/davinci.c
@@ -32,9 +32,9 @@
 #include <linux/io.h>
 #include <linux/gpio.h>
 
-#include <asm/arch/hardware.h>
-#include <asm/arch/memory.h>
-#include <asm/arch/gpio.h>
+#include <mach/hardware.h>
+#include <mach/memory.h>
+
 #include <asm/mach-types.h>
 
 #include "musb_core.h"

Re: [PATCH] musb: fix DaVinci build again

[Sergei Shtylyov]

Hello, I wrote:

>>>>> The last Kevin Hilman's patch seems to have been broken on its way 
>>>>> to Greg's
>>>>> tree and also probably clashed with yet another patch. Fix DaVinci 
>>>>> build again.
>>>>> Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
>>>>> ---
>>>>> Felipe, I have to ask you again to be more careful with patches...

>>>>    11 days have passed since this has benn sent out and I'm still 
>>>> not seeing this 2.6.28 *hot* fix in the MUSB git tree. Greg, could 
>>>> you queue this urgently, bypassing the MUSB tree?

>>>   I'll be happy if it gets combined with the following patch in your 
>>> tree:

>>> http://www.kernel.org/pub/linux/kernel/people/gregkh/gregkh-2.6/patches/usb/usb-musb-build-fixes-for-davinci.patch 

>>> fallout of which I'm fixing. Though it would've been better yet if 
>>> you just dropped that totally broken patch. I could've send a 
>>> replacment then...

>>   The patch I liked to above fixes nothing and only adds more 
>> breakage. Compare it with the original Kevin's patch:

>> http://marc.info/?l=linux-usb&m=122608723815504

>   I'm seeing Felipe returning to the list but still ignoring this thread 
> so far.
>   For how much longer we'll have to wait for this issue to be resolved? 
> 2.6.28 is 2 days away (at most).
>   Greg, what's the problem with dropping obviously broken patch?

    Greg, looks like that *2.6.28* (acyually even 2.6.27) patch is even in the 
wrong sub-series (past-2.6.28) of your tree, so there was no hope on the 
timely resolution from the very start... sigh. :-(

WBR, Sergei