[PATCH] DMA: shdma: fix a bad merge - remove free_irq()

[PATCH] DMA: shdma: fix a bad merge - remove free_irq()

[Guennadi Liakhovetski]

A bad merge resulted in a left-over free_irq() call. This patch removes it.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski+renesas@gmail.com>
---

Against today's dma-slave next branch.

 drivers/dma/sh/shdmac.c |    5 -----
 1 files changed, 0 insertions(+), 5 deletions(-)

diff --git a/drivers/dma/sh/shdmac.c b/drivers/dma/sh/shdmac.c
index c7faded..8f1683d 100644
--- a/drivers/dma/sh/shdmac.c
+++ b/drivers/dma/sh/shdmac.c
@@ -894,14 +894,9 @@ static int sh_dmae_remove(struct platform_device *pdev)
 {
 	struct sh_dmae_device *shdev = platform_get_drvdata(pdev);
 	struct dma_device *dma_dev = &shdev->shdma_dev.dma_dev;
-	struct resource *res;
-	int errirq = platform_get_irq(pdev, 0);
 
 	dma_async_device_unregister(dma_dev);
 
-	if (errirq > 0)
-		free_irq(errirq, shdev);
-
 	spin_lock_irq(&sh_dmae_lock);
 	list_del_rcu(&shdev->node);
 	spin_unlock_irq(&sh_dmae_lock);
-- 
1.7.2.5

Re: [PATCH] DMA: shdma: fix a bad merge - remove free_irq()

[Vinod Koul]

On Mon, Aug 26, 2013 at 02:27:51PM +0200, Guennadi Liakhovetski wrote:
> A bad merge resulted in a left-over free_irq() call. This patch removes it.
Applied, thanks

Can you also test this branch and see if anything is remaining

And can you take care then ARM patches in last serries are merged this week...

~Vinod

Re: [PATCH] DMA: shdma: fix a bad merge - remove free_irq()

[Guennadi Liakhovetski]

Hi Vinod,

On Mon, 26 Aug 2013, Vinod Koul wrote:

> On Mon, Aug 26, 2013 at 02:27:51PM +0200, Guennadi Liakhovetski wrote:
> > A bad merge resulted in a left-over free_irq() call. This patch removes it.
> Applied, thanks
> 
> Can you also test this branch and see if anything is remaining
> 
> And can you take care then ARM patches in last serries are merged this week...

Ok, I wanted to just test -next, but the latest is still of 22 Aug and 
that's too old. So, I took the current sh-mobile tree and merged your 
tree on top. There was 1 merge conflict in arch/arm/Makefile, but that's 
unrelated. After doing that I've discovered the following problems:

1. you applied a wrong version (v1) of my patch "DMA: shdma: support the 
new CHCLR register layout" - not v2, whose patchwork ID I provided:

> DMA: shdma: support the new CHCLR register layout
> pw ID: 2825593

I suggest to revert that commit and apply the correct one.

2. you missed this patch:

> DMA: shdma: move two macros to a header
> pw ID: 2831934

Please, pick it up too, even though it's just a cosmetic one.

3. Your merge:

commit ba1e06e3dff9a0bb0758d169e50cbb04a67a402c
Merge: daabb76 74cd26c
Author: Vinod Koul <vinod.koul@intel.com>
Date:   Sun Aug 25 13:22:06 2013 +0530

    Merge branch 'topic/sh' into next
    
    Conflicts:
        drivers/dma/sh/shdmac.c

dropped a "const" from a variable, which now generates a compiler warning. 
I'll send an incremental fix for that soon.

After fixing all the above the tree seems to be in a proper state. After 
your tree is fixed we can push ARM patches too.

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

Re: [PATCH] DMA: shdma: fix a bad merge - remove free_irq()

[Vinod Koul]

On Tue, Aug 27, 2013 at 11:21:00AM +0200, Guennadi Liakhovetski wrote:
> Hi Vinod,
> 
> On Mon, 26 Aug 2013, Vinod Koul wrote:
> 
> > On Mon, Aug 26, 2013 at 02:27:51PM +0200, Guennadi Liakhovetski wrote:
> > > A bad merge resulted in a left-over free_irq() call. This patch removes it.
> > Applied, thanks
> > 
> > Can you also test this branch and see if anything is remaining
> > 
> > And can you take care then ARM patches in last serries are merged this week...
> 
> Ok, I wanted to just test -next, but the latest is still of 22 Aug and 
> that's too old. So, I took the current sh-mobile tree and merged your 
> tree on top. There was 1 merge conflict in arch/arm/Makefile, but that's 
> unrelated. After doing that I've discovered the following problems:
> 
> 1. you applied a wrong version (v1) of my patch "DMA: shdma: support the 
> new CHCLR register layout" - not v2, whose patchwork ID I provided:
> 
> > DMA: shdma: support the new CHCLR register layout
> > pw ID: 2825593
I have redone the topic branch picking the v2 patch

> I suggest to revert that commit and apply the correct one.
> 
> 2. you missed this patch:
> 
> > DMA: shdma: move two macros to a header
> > pw ID: 2831934
> 
> Please, pick it up too, even though it's just a cosmetic one.
I ahve fixed that up in merge, pls verfiy. So need need of update patch which
you sent..

~Vinod

Re: [PATCH] DMA: shdma: fix a bad merge - remove free_irq()

[Guennadi Liakhovetski]

On Tue, 27 Aug 2013, Vinod Koul wrote:

> On Tue, Aug 27, 2013 at 11:21:00AM +0200, Guennadi Liakhovetski wrote:
> > Hi Vinod,
> > 
> > On Mon, 26 Aug 2013, Vinod Koul wrote:
> > 
> > > On Mon, Aug 26, 2013 at 02:27:51PM +0200, Guennadi Liakhovetski wrote:
> > > > A bad merge resulted in a left-over free_irq() call. This patch removes it.
> > > Applied, thanks
> > > 
> > > Can you also test this branch and see if anything is remaining
> > > 
> > > And can you take care then ARM patches in last serries are merged this week...
> > 
> > Ok, I wanted to just test -next, but the latest is still of 22 Aug and 
> > that's too old. So, I took the current sh-mobile tree and merged your 
> > tree on top. There was 1 merge conflict in arch/arm/Makefile, but that's 
> > unrelated. After doing that I've discovered the following problems:
> > 
> > 1. you applied a wrong version (v1) of my patch "DMA: shdma: support the 
> > new CHCLR register layout" - not v2, whose patchwork ID I provided:
> > 
> > > DMA: shdma: support the new CHCLR register layout
> > > pw ID: 2825593
> I have redone the topic branch picking the v2 patch
> 
> > I suggest to revert that commit and apply the correct one.
> > 
> > 2. you missed this patch:
> > 
> > > DMA: shdma: move two macros to a header
> > > pw ID: 2831934
> > 
> > Please, pick it up too, even though it's just a cosmetic one.
> I ahve fixed that up in merge, pls verfiy. So need need of update patch which
> you sent..

Ok, to me it looks like you did fix some patches in the topic/sh branch, 
but:

patch "DMA: shdma: move two macros to a header" is missing
patch "dma: use dev_get_platdata()" is missing (but ok, it's common for 
drivers/dma, not sh-specific)
patch "shdma: fixup sh_dmae_get_partial() calculation error" is missing

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/

Re: [PATCH] DMA: shdma: fix a bad merge - remove free_irq()

[Vinod Koul]

On Tue, Aug 27, 2013 at 04:46:51PM +0200, Guennadi Liakhovetski wrote:
> Ok, to me it looks like you did fix some patches in the topic/sh branch, 
> but:
> 
> patch "DMA: shdma: move two macros to a header" is missing
> patch "dma: use dev_get_platdata()" is missing (but ok, it's common for 
> drivers/dma, not sh-specific)
I havent applied that yet, its in my queue..
> patch "shdma: fixup sh_dmae_get_partial() calculation error" is missing
Can you resend these two...?

~Vinod

Re: [PATCH] DMA: shdma: fix a bad merge - remove free_irq()

[Guennadi Liakhovetski]

Hi Vinod,

On Wed, 28 Aug 2013, Vinod Koul wrote:

> On Tue, Aug 27, 2013 at 04:46:51PM +0200, Guennadi Liakhovetski wrote:
> > Ok, to me it looks like you did fix some patches in the topic/sh branch, 
> > but:
> > 
> > patch "DMA: shdma: move two macros to a header" is missing
> > patch "dma: use dev_get_platdata()" is missing (but ok, it's common for 
> > drivers/dma, not sh-specific)
> I havent applied that yet, its in my queue..
> > patch "shdma: fixup sh_dmae_get_partial() calculation error" is missing
> Can you resend these two...?

Ok, sorry, there was some confusion: yesterday no -next tree was 
available, so, I had to use the ARM sh-mobile tree with your slave-dma 
"next" branch merged on top. So, of the 3 above patches

1. "DMA: shdma: move two macros to a header"
2. "dma: use dev_get_platdata()"
3. "shdma: fixup sh_dmae_get_partial() calculation error"

#2 and #3 are already in "next" as 3c4d927 and d4adcc0 respectively, so, 
no need to resend. #1 is available here 
https://patchwork.kernel.org/patch/2831934/

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/