[12/11] ARM: OMAP: Fix relative includes for fpga.h

Commit Message

Benoit Cousson Nov. 12, 2012, 10:47 a.m. UTC

Hi Tony,

After rebasing on top of your omap-for-v3.8/tmp-merge, I realized that this patch is generating a build break due to removal of the fpga.h include from debug-leds.c.

arch/arm/plat-omap/debug-leds.c: In function 'fpga_probe':
arch/arm/plat-omap/debug-leds.c:114:9: error: 'H2P2_DBG_FPGA_SIZE' undeclared (first use in this function)
arch/arm/plat-omap/debug-leds.c:114:9: note: each undeclared identifier is reported only once for each function it appears in

It does not appear with the omap2plus_config but will appear if you enable the NEW_LEDS support that enable OMAP_DEBUG_LEDS.

I'm not sure why the size is still hard coded. In theory, assuming the resource is properly initialized, resource_size should be usable.
The patch below is fixing that.

Regards,
Benoit

---
From 81e1c3dddd1cd78b1f5a04983cf920b287764f11 Mon Sep 17 00:00:00 2001
From: Benoit Cousson <[e-mail]>
Date: Mon, 12 Nov 2012 11:44:03 +0100
Subject: [PATCH] ARM: OMAP: debug-leds: Use resource_size instead of hard coded macro

The debug-leds driver should not rely on hard coded macro for
the iomem size but use the resource size instead.

Signed-off-by: Benoit Cousson <[e-mail]>
---
 arch/arm/plat-omap/debug-leds.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

Comments

Tony Lindgren Nov. 12, 2012, 9:34 p.m. UTC | #1

* Benoit Cousson <[e-mail]> [121112 02:49]:
> Hi Tony,
> 
> After rebasing on top of your omap-for-v3.8/tmp-merge, I realized that this patch is generating a build break due to removal of the fpga.h include from debug-leds.c.
> 
> arch/arm/plat-omap/debug-leds.c: In function 'fpga_probe':
> arch/arm/plat-omap/debug-leds.c:114:9: error: 'H2P2_DBG_FPGA_SIZE' undeclared (first use in this function)
> arch/arm/plat-omap/debug-leds.c:114:9: note: each undeclared identifier is reported only once for each function it appears in
> 
> It does not appear with the omap2plus_config but will appear if you enable the NEW_LEDS support that enable OMAP_DEBUG_LEDS.
> 
> I'm not sure why the size is still hard coded. In theory, assuming the resource is properly initialized, resource_size should be usable.
> The patch below is fixing that.

Thanks looks like I missed that one. Applying into
omap-for-v3.8/cleanup-headers-prepare-multiplatform-v3.

Regards,

Tony

Patch

diff --git a/arch/arm/plat-omap/debug-leds.c b/arch/arm/plat-omap/debug-leds.c
index c43ea21..aa7ebc6 100644
--- a/arch/arm/plat-omap/debug-leds.c
+++ b/arch/arm/plat-omap/debug-leds.c
@@ -111,7 +111,7 @@  static int fpga_probe(struct platform_device *pdev)
 	if (!iomem)
 		return -ENODEV;
 
-	fpga = ioremap(iomem->start, H2P2_DBG_FPGA_SIZE);
+	fpga = ioremap(iomem->start, resource_size(iomem));
 	__raw_writew(0xff, &fpga->leds);
 
 	for (i = 0; i < ARRAY_SIZE(dbg_leds); i++) {