[PATCH] palm_bk3710: fix resource management (take 2)

[PATCH] palm_bk3710: fix resource management (take 2)

[Sergei Shtylyov]

The driver expected a *virtual* address in the IDE platform device's memory
resource and didn't request the memory region for the register block. Fix this
taking into account the fact that DaVinci SoC devices are fixed-mapped to the
virtual memory early and we can get their virtual addresses using IO_ADDRESS()
macro, not having to call ioremap()...

While at it, also do some cosmetic changes...

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>

---
Argh! I have finally run checkpatch.pl and it detected some whitespace damage.
So, here's take 2 of that stupid patch...

 drivers/ide/arm/palm_bk3710.c |   22 +++++++++++++---------
 1 files changed, 13 insertions(+), 9 deletions(-)

Index: linux-2.6/drivers/ide/arm/palm_bk3710.c
===================================================================
--- linux-2.6.orig/drivers/ide/arm/palm_bk3710.c
+++ linux-2.6/drivers/ide/arm/palm_bk3710.c
@@ -353,8 +353,8 @@ static int __devinit palm_bk3710_probe(s
 	struct clk *clkp;
 	struct resource *mem, *irq;
 	ide_hwif_t *hwif;
-	void __iomem *base;
-	int pribase, i;
+	unsigned long base;
+	int i;
 	hw_regs_t hw;
 	u8 idx[4] = { 0xff, 0xff, 0xff, 0xff };
 
@@ -374,22 +374,27 @@ static int __devinit palm_bk3710_probe(s
 		printk(KERN_ERR "failed to get memory region resource\n");
 		return -ENODEV;
 	}
+
 	irq = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
 	if (irq == NULL) {
 		printk(KERN_ERR "failed to get IRQ resource\n");
 		return -ENODEV;
 	}
 
-	base = (void *)mem->start;
+	if (request_mem_region(mem->start, mem->end - mem->start + 1,
+			       "palm_bk3710") == NULL) {
+		printk(KERN_ERR "failed to request memory region\n");
+		return -EBUSY;
+	}
+
+	base = IO_ADDRESS(mem->start);
 
 	/* Configure the Palm Chip controller */
-	palm_bk3710_chipinit(base);
+	palm_bk3710_chipinit((void __iomem *)base);
 
-	pribase = mem->start + IDE_PALM_ATA_PRI_REG_OFFSET;
 	for (i = 0; i < IDE_NR_PORTS - 2; i++)
-		hw.io_ports_array[i] = pribase + i;
-	hw.io_ports.ctl_addr = mem->start +
-			IDE_PALM_ATA_PRI_CTL_OFFSET;
+ 		hw.io_ports_array[i] = base + IDE_PALM_ATA_PRI_REG_OFFSET + i;
+	hw.io_ports.ctl_addr = base + IDE_PALM_ATA_PRI_CTL_OFFSET;
 	hw.irq = irq->start;
 	hw.chipset = ide_palm3710;
 
@@ -434,4 +439,3 @@ static int __init palm_bk3710_init(void)
 
 module_init(palm_bk3710_init);
 MODULE_LICENSE("GPL");
-

[PATCH] palm_bk3710: fix resource management (take 2)

[Sergei Shtylyov]

The driver expected a *virtual* address in the IDE platform device's memory
resource and didn't request the memory region for the register block. Fix this
taking into account the fact that DaVinci SoC devices are fixed-mapped to the
virtual memory early and we can get their virtual addresses using IO_ADDRESS()
macro, not having to call ioremap()...

While at it, also do some cosmetic changes...

Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>

---
Argh! I have finally run checkpatch.pl and it detected some whitespace damage.
And of course, I then re-sent it without fixing that!
So, here's take 2 of that stupid patch for real now. Sorry for the mess... :-<

 drivers/ide/arm/palm_bk3710.c |   22 +++++++++++++---------
 1 files changed, 13 insertions(+), 9 deletions(-)

Index: linux-2.6/drivers/ide/arm/palm_bk3710.c
===================================================================
--- linux-2.6.orig/drivers/ide/arm/palm_bk3710.c
+++ linux-2.6/drivers/ide/arm/palm_bk3710.c
@@ -353,8 +353,8 @@ static int __devinit palm_bk3710_probe(s
 	struct clk *clkp;
 	struct resource *mem, *irq;
 	ide_hwif_t *hwif;
-	void __iomem *base;
-	int pribase, i;
+	unsigned long base;
+	int i;
 	hw_regs_t hw;
 	u8 idx[4] = { 0xff, 0xff, 0xff, 0xff };
 
@@ -374,22 +374,27 @@ static int __devinit palm_bk3710_probe(s
 		printk(KERN_ERR "failed to get memory region resource\n");
 		return -ENODEV;
 	}
+
 	irq = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
 	if (irq == NULL) {
 		printk(KERN_ERR "failed to get IRQ resource\n");
 		return -ENODEV;
 	}
 
-	base = (void *)mem->start;
+	if (request_mem_region(mem->start, mem->end - mem->start + 1,
+			       "palm_bk3710") == NULL) {
+		printk(KERN_ERR "failed to request memory region\n");
+		return -EBUSY;
+	}
+
+	base = IO_ADDRESS(mem->start);
 
 	/* Configure the Palm Chip controller */
-	palm_bk3710_chipinit(base);
+	palm_bk3710_chipinit((void __iomem *)base);
 
-	pribase = mem->start + IDE_PALM_ATA_PRI_REG_OFFSET;
 	for (i = 0; i < IDE_NR_PORTS - 2; i++)
-		hw.io_ports_array[i] = pribase + i;
-	hw.io_ports.ctl_addr = mem->start +
-			IDE_PALM_ATA_PRI_CTL_OFFSET;
+		hw.io_ports_array[i] = base + IDE_PALM_ATA_PRI_REG_OFFSET + i;
+	hw.io_ports.ctl_addr = base + IDE_PALM_ATA_PRI_CTL_OFFSET;
 	hw.irq = irq->start;
 	hw.chipset = ide_palm3710;
 
@@ -434,4 +439,3 @@ static int __init palm_bk3710_init(void)
 
 module_init(palm_bk3710_init);
 MODULE_LICENSE("GPL");
-

Re: [PATCH] palm_bk3710: fix resource management (take 2)

[Bartlomiej Zolnierkiewicz]

On Monday 16 June 2008, Sergei Shtylyov wrote:
> The driver expected a *virtual* address in the IDE platform device's memory
> resource and didn't request the memory region for the register block. Fix this
> taking into account the fact that DaVinci SoC devices are fixed-mapped to the
> virtual memory early and we can get their virtual addresses using IO_ADDRESS()
> macro, not having to call ioremap()...
> 
> While at it, also do some cosmetic changes...
> 
> Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
> 
> ---
> Argh! I have finally run checkpatch.pl and it detected some whitespace damage.
> And of course, I then re-sent it without fixing that!
> So, here's take 2 of that stupid patch for real now. Sorry for the mess... :-<

I've already fixed it while applying "take 1" but I was too lazy to
mention it (not that single " " removal is worth mentioning)... :)

[ However please integrate checkpatch.pl into your workflow... ]

Re: [PATCH] palm_bk3710: fix resource management (take 2)

[Sergei Shtylyov]

Hello.

Bartlomiej Zolnierkiewicz wrote:
> On Monday 16 June 2008, Sergei Shtylyov wrote:
>   
>> The driver expected a *virtual* address in the IDE platform device's memory
>> resource and didn't request the memory region for the register block. Fix this
>> taking into account the fact that DaVinci SoC devices are fixed-mapped to the
>> virtual memory early and we can get their virtual addresses using IO_ADDRESS()
>> macro, not having to call ioremap()...
>>
>> While at it, also do some cosmetic changes...
>>
>> Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com>
>>
>> ---
>> Argh! I have finally run checkpatch.pl and it detected some whitespace damage.
>> And of course, I then re-sent it without fixing that!
>> So, here's take 2 of that stupid patch for real now. Sorry for the mess... :-<
>>     
>
> I've already fixed it while applying "take 1" but I was too lazy to
> mention it (not that single " " removal is worth mentioning)... :)
>   

   I do that kind of mistakes in reject resolution very rarely. This 
just wasn't my day it seems... :-)

> [ However please integrate checkpatch.pl into your workflow... ]
>
>   
   The forgotten signoff is surely my problem, so indeed I should... if 
I don't forget. ;-)

WBR. Sergei