Skip to contentSkip to conversation

64bit resources start end value fix

PATCH In review Vivek Goyal 4 participants 13 messages Jun 21, 2006 – Jun 24, 2006

Subsystems (11)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 64bit resources start end value fix

Hi Greg,

While changing 64bit kconfig options to CONFIG_RESOURCES_64BIT, I forgot to update the values of start and end fields in ioport_resource and iomem_resource.

Following patch applies on top of your reworked 64 bit patches and is based on Andrew Morton's patch. Please apply.

http://marc.theaimsgroup.com/?l=linux-mm-commits&m=115087406130723&w=2

Thanks
Vivek

o Update start and end fields for 32bit and 64bit resources.

---

 linux-2.6.17-1M-vivek/kernel/resource.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)
kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +3 -3
diff -puN kernel/resource.c~64bit-resources-start-end-value-fix kernel/resource.c
@@ -23,7 +23,7 @@
 
 struct resource ioport_resource = {
 	.name	= "PCI IO",
-	.start	= 0x0000,
+	.start	= 0,
 	.end	= IO_SPACE_LIMIT,
 	.flags	= IORESOURCE_IO,
 };
@@ -32,8 +32,8 @@ EXPORT_SYMBOL(ioport_resource);
 
 struct resource iomem_resource = {
 	.name	= "PCI mem",
-	.start	= 0UL,
-	.end	= ~0UL,
+	.start	= 0,
+	.end	= -1,
 	.flags	= IORESOURCE_MEM,
 };
 

_

#1 lore raw
Quoted 37 lines · On Wed, 21 Jun 2006 13:29:03 -0400 Vivek Goyal <[email protected]> wrote:

Hi Greg,

While changing 64bit kconfig options to CONFIG_RESOURCES_64BIT, I forgot to update the values of start and end fields in ioport_resource and iomem_resource.

Following patch applies on top of your reworked 64 bit patches and is based on Andrew Morton's patch. Please apply.

http://marc.theaimsgroup.com/?l=linux-mm-commits&m=115087406130723&w=2

Thanks
Vivek

o Update start and end fields for 32bit and 64bit resources.

---

 linux-2.6.17-1M-vivek/kernel/resource.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff -puN kernel/resource.c~64bit-resources-start-end-value-fix kernel/resource.c

kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +1 -1
@@ -23,7 +23,7 @@
 
 struct resource ioport_resource = {
 	.name	= "PCI IO",
-	.start	= 0x0000,
+	.start	= 0,
 	.end	= IO_SPACE_LIMIT,
 	.flags	= IORESOURCE_IO,
 };

kernel/resource.c EXPORT_SYMBOL(ioport_resource); · line 39

Quoted 10 lines
kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +2 -2
@@ -32,8 +32,8 @@ EXPORT_SYMBOL(ioport_resource);
 
 struct resource iomem_resource = {
 	.name	= "PCI mem",
-	.start	= 0UL,
-	.end	= ~0UL,
+	.start	= 0,
+	.end	= -1,
 	.flags	= IORESOURCE_MEM,
 };
 

Confused. This patch won't apply. It will apply with `patch -R', and if you do that you'll break iomem_reosurce.end by setting it to 0x00000000ffffffff.

I don't think any additional changes are needed here.

#2 lore raw
Quoted 55 lines · Andrew Morton
Quoted 37 lines · On Wed, 21 Jun 2006 13:29:03 -0400 Vivek Goyal <[email protected]> wrote:

Hi Greg,

While changing 64bit kconfig options to CONFIG_RESOURCES_64BIT, I forgot to update the values of start and end fields in ioport_resource and iomem_resource.

Following patch applies on top of your reworked 64 bit patches and is based on Andrew Morton's patch. Please apply.

http://marc.theaimsgroup.com/?l=linux-mm-commits&m=115087406130723&w=2

Thanks
Vivek

o Update start and end fields for 32bit and 64bit resources.

---

 linux-2.6.17-1M-vivek/kernel/resource.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff -puN kernel/resource.c~64bit-resources-start-end-value-fix kernel/resource.c

kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +1 -1
@@ -23,7 +23,7 @@
 
 struct resource ioport_resource = {
 	.name	= "PCI IO",
-	.start	= 0x0000,
+	.start	= 0,
 	.end	= IO_SPACE_LIMIT,
 	.flags	= IORESOURCE_IO,
 };

kernel/resource.c EXPORT_SYMBOL(ioport_resource); · line 39

Quoted 10 lines
kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +2 -2
@@ -32,8 +32,8 @@ EXPORT_SYMBOL(ioport_resource);
 
 struct resource iomem_resource = {
 	.name	= "PCI mem",
-	.start	= 0UL,
-	.end	= ~0UL,
+	.start	= 0,
+	.end	= -1,
 	.flags	= IORESOURCE_MEM,
 };
 

Confused. This patch won't apply. It will apply with `patch -R', and if you do that you'll break iomem_reosurce.end by setting it to 0x00000000ffffffff.

Quoted 1 line

I don't think any additional changes are needed here.

Andrew, you don't have to apply this patch. It is supposed to be picked by Greg.

There seems to be some confusion. Just few days back Greg consolidated and re-organized all the 64bit resources patches and posted on LKML for review.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115015916118671&w=2

There were few review comments regarding kconfig options. I reworked the patch and CONFING_RESOURCES_32BIT was changed to CONFIG_RESOURCES_64BIT.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115072559700302&w=2

Now Greg's tree and your tree are not exact replica when it comes to 64bit resource patches. Hence this patch is supposed to be picked by Greg to make sure things are not broken in his tree.

Thanks
Vivek

#3 lore raw
Quoted 71 lines · Vivek Goyal
Quoted 55 lines · Andrew Morton
Quoted 37 lines · On Wed, 21 Jun 2006 13:29:03 -0400 Vivek Goyal <[email protected]> wrote:

Hi Greg,

While changing 64bit kconfig options to CONFIG_RESOURCES_64BIT, I forgot to update the values of start and end fields in ioport_resource and iomem_resource.

Following patch applies on top of your reworked 64 bit patches and is based on Andrew Morton's patch. Please apply.

http://marc.theaimsgroup.com/?l=linux-mm-commits&m=115087406130723&w=2

Thanks
Vivek

o Update start and end fields for 32bit and 64bit resources.

---

 linux-2.6.17-1M-vivek/kernel/resource.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff -puN kernel/resource.c~64bit-resources-start-end-value-fix kernel/resource.c

kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +1 -1
@@ -23,7 +23,7 @@
 
 struct resource ioport_resource = {
 	.name	= "PCI IO",
-	.start	= 0x0000,
+	.start	= 0,
 	.end	= IO_SPACE_LIMIT,
 	.flags	= IORESOURCE_IO,
 };

kernel/resource.c EXPORT_SYMBOL(ioport_resource); · line 39

Quoted 10 lines
kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +2 -2
@@ -32,8 +32,8 @@ EXPORT_SYMBOL(ioport_resource);
 
 struct resource iomem_resource = {
 	.name	= "PCI mem",
-	.start	= 0UL,
-	.end	= ~0UL,
+	.start	= 0,
+	.end	= -1,
 	.flags	= IORESOURCE_MEM,
 };
 

Confused. This patch won't apply. It will apply with `patch -R', and if you do that you'll break iomem_reosurce.end by setting it to 0x00000000ffffffff.

Quoted 1 line

I don't think any additional changes are needed here.

Andrew, you don't have to apply this patch. It is supposed to be picked by Greg.

There seems to be some confusion. Just few days back Greg consolidated and re-organized all the 64bit resources patches and posted on LKML for review.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115015916118671&w=2

There were few review comments regarding kconfig options. I reworked the patch and CONFING_RESOURCES_32BIT was changed to CONFIG_RESOURCES_64BIT.

Quoted 5 lines

http://marc.theaimsgroup.com/?l=linux-kernel&m=115072559700302&w=2

Now Greg's tree and your tree are not exact replica when it comes to 64bit resource patches. Hence this patch is supposed to be picked by Greg to make sure things are not broken in his tree.

It still breaks things as Andrew pointed out. .end should not be set to -1.

thanks,

greg k-h

#4 lore raw
Quoted 72 lines · Greg KH
Quoted 71 lines · Vivek Goyal
Quoted 55 lines · Andrew Morton
Quoted 37 lines · On Wed, 21 Jun 2006 13:29:03 -0400 Vivek Goyal <[email protected]> wrote:

Hi Greg,

While changing 64bit kconfig options to CONFIG_RESOURCES_64BIT, I forgot to update the values of start and end fields in ioport_resource and iomem_resource.

Following patch applies on top of your reworked 64 bit patches and is based on Andrew Morton's patch. Please apply.

http://marc.theaimsgroup.com/?l=linux-mm-commits&m=115087406130723&w=2

Thanks
Vivek

o Update start and end fields for 32bit and 64bit resources.

---

 linux-2.6.17-1M-vivek/kernel/resource.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff -puN kernel/resource.c~64bit-resources-start-end-value-fix kernel/resource.c

kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +1 -1
@@ -23,7 +23,7 @@
 
 struct resource ioport_resource = {
 	.name	= "PCI IO",
-	.start	= 0x0000,
+	.start	= 0,
 	.end	= IO_SPACE_LIMIT,
 	.flags	= IORESOURCE_IO,
 };

kernel/resource.c EXPORT_SYMBOL(ioport_resource); · line 39

Quoted 10 lines
kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +2 -2
@@ -32,8 +32,8 @@ EXPORT_SYMBOL(ioport_resource);
 
 struct resource iomem_resource = {
 	.name	= "PCI mem",
-	.start	= 0UL,
-	.end	= ~0UL,
+	.start	= 0,
+	.end	= -1,
 	.flags	= IORESOURCE_MEM,
 };
 

Confused. This patch won't apply. It will apply with `patch -R', and if you do that you'll break iomem_reosurce.end by setting it to 0x00000000ffffffff.

Quoted 1 line

I don't think any additional changes are needed here.

Andrew, you don't have to apply this patch. It is supposed to be picked by Greg.

There seems to be some confusion. Just few days back Greg consolidated and re-organized all the 64bit resources patches and posted on LKML for review.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115015916118671&w=2

There were few review comments regarding kconfig options. I reworked the patch and CONFING_RESOURCES_32BIT was changed to CONFIG_RESOURCES_64BIT.

Quoted 8 lines
Quoted 5 lines

http://marc.theaimsgroup.com/?l=linux-kernel&m=115072559700302&w=2

Now Greg's tree and your tree are not exact replica when it comes to 64bit resource patches. Hence this patch is supposed to be picked by Greg to make sure things are not broken in his tree.

It still breaks things as Andrew pointed out. .end should not be set to -1.

I think Andrew mentioned that it breaks things if applied on -mm with -R. Because it will set .end to ~0UL instead of -1

Is .end = -1 wrong? Won't it effectively be .end = 0xffffffffffffffff for 64 bit resources?

Thanks
Vivek

#6 lore raw
In reply to Vivek Goyal
Quoted 31 lines · On Wed, 21 Jun 2006 16:41:21 -0400 Vivek Goyal <[email protected]> wrote:
Quoted 16 lines

kernel/resource.c EXPORT_SYMBOL(ioport_resource); · line 39

Quoted 11 lines
kernel/resource.c renamed from kernel/resource.c~64bit-resources-start-end-value-fix +2 -2
@@ -32,8 +32,8 @@ EXPORT_SYMBOL(ioport_resource);
 
 struct resource iomem_resource = {
 	.name	= "PCI mem",
-	.start	= 0UL,
-	.end	= ~0UL,
+	.start	= 0,
+	.end	= -1,
 	.flags	= IORESOURCE_MEM,
 };
 

Confused. This patch won't apply. It will apply with `patch -R', and if you do that you'll break iomem_reosurce.end by setting it to 0x00000000ffffffff.

Quoted 1 line

I don't think any additional changes are needed here.

Andrew, you don't have to apply this patch. It is supposed to be picked by Greg.

There seems to be some confusion. Just few days back Greg consolidated and re-organized all the 64bit resources patches and posted on LKML for review.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115015916118671&w=2

There were few review comments regarding kconfig options. I reworked the patch and CONFING_RESOURCES_32BIT was changed to CONFIG_RESOURCES_64BIT.

Quoted 5 lines

http://marc.theaimsgroup.com/?l=linux-kernel&m=115072559700302&w=2

Now Greg's tree and your tree are not exact replica when it comes to 64bit resource patches. Hence this patch is supposed to be picked by Greg to make sure things are not broken in his tree.

I'm working against Greg's tree, always...

#5 lore raw
In reply to Greg KH
Quoted 19 lines · On Wed, 21 Jun 2006 13:44:15 -0700 Greg KH <[email protected]> wrote:
Quoted 19 lines
Quoted 5 lines

Confused. This patch won't apply. It will apply with `patch -R', and if you do that you'll break iomem_reosurce.end by setting it to 0x00000000ffffffff.

I don't think any additional changes are needed here.

Andrew, you don't have to apply this patch. It is supposed to be picked by Greg.

There seems to be some confusion. Just few days back Greg consolidated and re-organized all the 64bit resources patches and posted on LKML for review.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115015916118671&w=2

There were few review comments regarding kconfig options. I reworked the patch and CONFING_RESOURCES_32BIT was changed to CONFIG_RESOURCES_64BIT.

Quoted 8 lines
Quoted 5 lines

http://marc.theaimsgroup.com/?l=linux-kernel&m=115072559700302&w=2

Now Greg's tree and your tree are not exact replica when it comes to 64bit resource patches. Hence this patch is supposed to be picked by Greg to make sure things are not broken in his tree.

It still breaks things as Andrew pointed out. .end should not be set to -1.

No, -1 is OK.

As it turns out,

	unsigned long long x = ~0UL;

sets `x' to 0xffffffffffffffff which was totally not what I expected.

But -1 works, and the patch I have against your tree is OK.

Could someone please fix Andy Isaacson <[email protected]>'s bug, btw?

#7 lore raw
In reply to Andrew Morton
Quoted 14 lines · Andrew Morton
Quoted 14 lines

Andrew, you don't have to apply this patch. It is supposed to be picked by Greg.

There seems to be some confusion. Just few days back Greg consolidated and re-organized all the 64bit resources patches and posted on LKML for review.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115015916118671&w=2

There were few review comments regarding kconfig options. I reworked the patch and CONFING_RESOURCES_32BIT was changed to CONFIG_RESOURCES_64BIT.

Quoted 7 lines
Quoted 5 lines

http://marc.theaimsgroup.com/?l=linux-kernel&m=115072559700302&w=2

Now Greg's tree and your tree are not exact replica when it comes to 64bit resource patches. Hence this patch is supposed to be picked by Greg to make sure things are not broken in his tree.

I'm working against Greg's tree, always...

I am sorry. That's a mistake on my part. I misunderstood it.

Can you please include the attached patch. Now RESOURCES_32BIT has been dropped and RESOURCES_64BIT has been introduced based on the review feedback.

By default RESOURCES_64BIT is not selected for 32bit platform and user has got the option to enable it. For 64BIT kernels, RESOURCES_64BIT is always set.

Thanks
Vivek

o Based on review comments, removed CONFIG_RESOURCE_32BIT and added CONFIG_RESOURCE_64BIT in arch independent manner.

---

 arch/arm/Kconfig         |    7 -------
 arch/arm26/Kconfig       |    7 -------
 arch/cris/Kconfig        |    7 -------
 arch/frv/Kconfig         |    7 -------
 arch/i386/Kconfig        |    8 +-------
 arch/i386/kernel/setup.c |    2 +-
 arch/m32r/Kconfig        |    7 -------
 arch/m68k/Kconfig        |    7 -------
 arch/m68knommu/Kconfig   |    7 -------
 arch/mips/Kconfig        |    8 --------
 arch/parisc/Kconfig      |    8 --------
 arch/powerpc/Kconfig     |    8 --------
 arch/ppc/Kconfig         |    7 -------
 arch/s390/Kconfig        |    8 --------
 arch/sh/Kconfig          |    7 -------
 arch/sparc/Kconfig       |    7 -------
 arch/v850/Kconfig        |    7 -------
 arch/xtensa/Kconfig      |    7 -------
 include/linux/types.h    |    6 +++---
 mm/Kconfig               |    6 ++++++
 20 files changed, 11 insertions(+), 127 deletions(-)
mm/Kconfig renamed from mm/Kconfig~64bit-resources-modify-kconfig-options +6 -0
diff -puN mm/Kconfig~64bit-resources-modify-kconfig-options mm/Kconfig
@@ -208,3 +208,9 @@ config READAHEAD_SMOOTH_AGING
 	def_bool n if NUMA
 	default y if !NUMA
 	depends on ADAPTIVE_READAHEAD
+
+config RESOURCES_64BIT
+	bool "64 bit Memory and IO resources (EXPERIMENTAL)" if (!64BIT && EXPERIMENTAL)
+	default 64BIT
+	help
+	  This option allows memory and IO resources to be 64 bit.
include/linux/types.h renamed from include/linux/types.h~64bit-resources-modify-kconfig-options +3 -3
diff -puN include/linux/types.h~64bit-resources-modify-kconfig-options include/linux/types.h
@@ -178,10 +178,10 @@ typedef __u64 __bitwise __be64;
 #ifdef __KERNEL__
 typedef unsigned __bitwise__ gfp_t;
 
-#ifdef CONFIG_RESOURCES_32BIT
-typedef __u32 resource_size_t;
+#ifdef CONFIG_RESOURCES_64BIT
+typedef u64 resource_size_t;
 #else
-typedef __u64 resource_size_t;
+typedef u32 resource_size_t;
 #endif
 
 #endif	/* __KERNEL__ */
arch/i386/Kconfig renamed from arch/i386/Kconfig~64bit-resources-modify-kconfig-options +1 -7
diff -puN arch/i386/Kconfig~64bit-resources-modify-kconfig-options arch/i386/Kconfig
@@ -528,6 +528,7 @@ config X86_PAE
 	bool
 	depends on HIGHMEM64G
 	default y
+	select RESOURCES_64BIT
 
 # Common NUMA Features
 config NUMA
@@ -774,13 +775,6 @@ config PHYSICAL_START
 
 	  Don't change this unless you know what you are doing.
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL && !X86_PAE
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 config HOTPLUG_CPU
 	bool "Support for hot-pluggable CPUs (EXPERIMENTAL)"
 	depends on SMP && HOTPLUG && EXPERIMENTAL && !X86_VOYAGER
arch/arm/Kconfig renamed from arch/arm/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/arm/Kconfig~64bit-resources-modify-kconfig-options arch/arm/Kconfig
@@ -531,13 +531,6 @@ config NODES_SHIFT
 
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 config LEDS
 	bool "Timer and CPU usage LEDs"
 	depends on ARCH_CDB89712 || ARCH_CO285 || ARCH_EBSA110 || \
arch/arm26/Kconfig renamed from arch/arm26/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/arm26/Kconfig~64bit-resources-modify-kconfig-options arch/arm26/Kconfig
@@ -187,13 +187,6 @@ config CMDLINE
 
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 endmenu
 
 source "net/Kconfig"
arch/cris/Kconfig renamed from arch/cris/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/cris/Kconfig~64bit-resources-modify-kconfig-options arch/cris/Kconfig
@@ -84,13 +84,6 @@ config PREEMPT
 
 source mm/Kconfig
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 endmenu
 
 menu "Hardware setup"
arch/frv/Kconfig renamed from arch/frv/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/frv/Kconfig~64bit-resources-modify-kconfig-options arch/frv/Kconfig
@@ -80,13 +80,6 @@ config HIGHPTE
 
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 choice
 	prompt "uClinux kernel load address"
 	depends on !MMU
arch/m32r/Kconfig renamed from arch/m32r/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/m32r/Kconfig~64bit-resources-modify-kconfig-options arch/m32r/Kconfig
@@ -188,13 +188,6 @@ config ARCH_DISCONTIGMEM_ENABLE
 
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 config IRAM_START
 	hex "Internal memory start address (hex)"
 	default "00f00000" if !CHIP_M32104
arch/m68k/Kconfig renamed from arch/m68k/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/m68k/Kconfig~64bit-resources-modify-kconfig-options arch/m68k/Kconfig
@@ -368,13 +368,6 @@ config 060_WRITETHROUGH
 
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 endmenu
 
 menu "General setup"
arch/m68knommu/Kconfig renamed from arch/m68knommu/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/m68knommu/Kconfig~64bit-resources-modify-kconfig-options arch/m68knommu/Kconfig
@@ -605,13 +605,6 @@ endchoice
 
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 endmenu
 
 config ISA_DMA_API
arch/mips/Kconfig renamed from arch/mips/Kconfig~64bit-resources-modify-kconfig-options +0 -8
diff -puN arch/mips/Kconfig~64bit-resources-modify-kconfig-options arch/mips/Kconfig
@@ -1789,14 +1789,6 @@ config HZ
 
 source "kernel/Kconfig.preempt"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	depends on 32BIT
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 config RTC_DS1742
 	bool "DS1742 BRAM/RTC support"
 	depends on TOSHIBA_JMR3927 || TOSHIBA_RBTX4927
arch/parisc/Kconfig renamed from arch/parisc/Kconfig~64bit-resources-modify-kconfig-options +0 -8
diff -puN arch/parisc/Kconfig~64bit-resources-modify-kconfig-options arch/parisc/Kconfig
@@ -221,14 +221,6 @@ source "kernel/Kconfig.preempt"
 source "kernel/Kconfig.hz"
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	depends on !64BIT
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 config COMPAT
 	def_bool y
 	depends on 64BIT
arch/powerpc/Kconfig renamed from arch/powerpc/Kconfig~64bit-resources-modify-kconfig-options +0 -8
diff -puN arch/powerpc/Kconfig~64bit-resources-modify-kconfig-options arch/powerpc/Kconfig
@@ -634,14 +634,6 @@ config CRASH_DUMP
 
 	  Don't change this unless you know what you are doing.
 
-config RESOURCES_32BIT
-        bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-        depends on EXPERIMENTAL
-	depends on PPC32
-        help
-          By default resources are 64 bit. This option allows memory and IO
-          resources to be 32 bit to optimize code size.
-
 config EMBEDDEDBOOT
 	bool
 	depends on 8xx || 8260
arch/ppc/Kconfig renamed from arch/ppc/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/ppc/Kconfig~64bit-resources-modify-kconfig-options arch/ppc/Kconfig
@@ -957,13 +957,6 @@ source kernel/Kconfig.hz
 source kernel/Kconfig.preempt
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 source "fs/Kconfig.binfmt"
 
 config PREP_RESIDUAL
arch/s390/Kconfig renamed from arch/s390/Kconfig~64bit-resources-modify-kconfig-options +0 -8
diff -puN arch/s390/Kconfig~64bit-resources-modify-kconfig-options arch/s390/Kconfig
@@ -218,14 +218,6 @@ config WARN_STACK_SIZE
 
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	depends on !64BIT
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 comment "I/O subsystem configuration"
 
 config MACHCHK_WARNING
arch/sh/Kconfig renamed from arch/sh/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/sh/Kconfig~64bit-resources-modify-kconfig-options arch/sh/Kconfig
@@ -532,13 +532,6 @@ config NODES_SHIFT
 	default "1"
 	depends on NEED_MULTIPLE_NODES
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 endmenu
 
 menu "Boot options"
arch/sparc/Kconfig renamed from arch/sparc/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/sparc/Kconfig~64bit-resources-modify-kconfig-options arch/sparc/Kconfig
@@ -67,13 +67,6 @@ config SPARC32
 	  maintains both the SPARC32 and SPARC64 ports; its web page is
 	  available at <http://www.ultralinux.org/>.
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 # Global things across all Sun machines.
 config ISA
 	bool
arch/v850/Kconfig renamed from arch/v850/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/v850/Kconfig~64bit-resources-modify-kconfig-options arch/v850/Kconfig
@@ -235,13 +235,6 @@ menu "Processor type and features"
 
 source "mm/Kconfig"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 endmenu
 
 
arch/xtensa/Kconfig renamed from arch/xtensa/Kconfig~64bit-resources-modify-kconfig-options +0 -7
diff -puN arch/xtensa/Kconfig~64bit-resources-modify-kconfig-options arch/xtensa/Kconfig
@@ -95,13 +95,6 @@ config MATH_EMULATION
 config HIGHMEM
 	bool "High memory support"
 
-config RESOURCES_32BIT
-	bool "32 bit Memory and IO resources (EXPERIMENTAL)"
-	depends on EXPERIMENTAL
-	help
-	  By default resources are 64 bit. This option allows memory and IO
-	  resources to be 32 bit to optimize code size.
-
 endmenu
 
 menu "Platform options"
arch/i386/kernel/setup.c renamed from arch/i386/kernel/setup.c~64bit-resources-modify-kconfig-options +1 -1
diff -puN arch/i386/kernel/setup.c~64bit-resources-modify-kconfig-options arch/i386/kernel/setup.c
@@ -1332,7 +1332,7 @@ legacy_init_iomem_resources(struct resou
 	probe_roms();
 	for (i = 0; i < e820.nr_map; i++) {
 		struct resource *res;
-#ifdef CONFIG_RESOURCES_32BIT
+#ifndef CONFIG_RESOURCES_64BIT
 		if (e820.map[i].addr + e820.map[i].size > 0x100000000ULL)
 			continue;
 #endif

_

#8 lore raw
Quoted 17 lines · On Wed, 21 Jun 2006 19:15:52 -0400 Vivek Goyal <[email protected]> wrote:
Quoted 16 lines · Andrew Morton
Quoted 16 lines

Andrew, you don't have to apply this patch. It is supposed to be picked by Greg.

There seems to be some confusion. Just few days back Greg consolidated and re-organized all the 64bit resources patches and posted on LKML for review.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115015916118671&w=2

There were few review comments regarding kconfig options. I reworked the patch and CONFING_RESOURCES_32BIT was changed to CONFIG_RESOURCES_64BIT.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115072559700302&w=2

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

Now Greg's tree and your tree are not exact replica when it comes to 64bit resource patches. Hence this patch is supposed to be picked by Greg to make sure things are not broken in his tree.

I'm working against Greg's tree, always...

I am sorry. That's a mistake on my part. I misunderstood it.

Oh.

Quoted 1 line

Can you please include the attached patch.

Hopefully I'll pick it up from
http://www.kernel.org/pub/linux/kernel/people/gregkh/gregkh-2.6/gregkh-03-pci
later today?

#9 lore raw
Quoted 28 lines · Andrew Morton
Quoted 17 lines · On Wed, 21 Jun 2006 19:15:52 -0400 Vivek Goyal <[email protected]> wrote:
Quoted 16 lines · Andrew Morton
Quoted 16 lines

Andrew, you don't have to apply this patch. It is supposed to be picked by Greg.

There seems to be some confusion. Just few days back Greg consolidated and re-organized all the 64bit resources patches and posted on LKML for review.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115015916118671&w=2

There were few review comments regarding kconfig options. I reworked the patch and CONFING_RESOURCES_32BIT was changed to CONFIG_RESOURCES_64BIT.

http://marc.theaimsgroup.com/?l=linux-kernel&m=115072559700302&w=2

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

Now Greg's tree and your tree are not exact replica when it comes to 64bit resource patches. Hence this patch is supposed to be picked by Greg to make sure things are not broken in his tree.

I'm working against Greg's tree, always...

I am sorry. That's a mistake on my part. I misunderstood it.

Quoted 7 lines

Oh.

Quoted 1 line

Can you please include the attached patch.

Hopefully I'll pick it up from
http://www.kernel.org/pub/linux/kernel/people/gregkh/gregkh-2.6/gregkh-03-pci
later today?

It will be tomorrow, sorry, catching up on other merge stuff...

thanks,

greg k-h

#10 lore raw
In reply to Vivek Goyal
Quoted 10 lines · Vivek Goyal

Hi Greg,

While changing 64bit kconfig options to CONFIG_RESOURCES_64BIT, I forgot to update the values of start and end fields in ioport_resource and iomem_resource.

Following patch applies on top of your reworked 64 bit patches and is based on Andrew Morton's patch. Please apply.

http://marc.theaimsgroup.com/?l=linux-mm-commits&m=115087406130723&w=2

Ok, I think I have this finally all straigned out. Can you look at my quilt tree to verify that I've tweaked everything properly based on this, and the other cleanup patches you and Andrew have been sending me recently?

thanks,

greg k-h

#11 lore raw
Quoted 10 lines · Greg KH
Quoted 9 lines · Vivek Goyal

Hi Greg,

While changing 64bit kconfig options to CONFIG_RESOURCES_64BIT, I forgot to update the values of start and end fields in ioport_resource and iomem_resource.

Following patch applies on top of your reworked 64 bit patches and is based on Andrew Morton's patch. Please apply.

Quoted 7 lines
Quoted 1 line

http://marc.theaimsgroup.com/?l=linux-mm-commits&m=115087406130723&w=2

Ok, I think I have this finally all straigned out. Can you look at my quilt tree to verify that I've tweaked everything properly based on this, and the other cleanup patches you and Andrew have been sending me recently?

I looked into following tree.

http://kernel.org/git/?p=linux/kernel/git/gregkh/patches.git;a=summary

Everything seems to be fine except that following i2c patch from andrew seems to be missing.

http://marc.theaimsgroup.com/?l=linux-mm-commits&m=115086650916817&w=2

Thanks
Vivek