BCM47xx patches
Subsystems (2)
cover BCM47xx patches
The following patches are a resent of patches sent early this year in order to support the BCM47xx chips:
[MIPS] BCM47xx: Add platform specific PCI code [MIPS] WGT634U: Add machine detection message [MIPS] Remove references to BCM947XX [MIPS] BCM47xx: Use the new SSB GPIO API [MIPS] Add WGT634U reset button support [MIPS] Scan PCI busses when they are registered
The first one fixes a build failure of the BCM47xx sub-architecture that has been introduced sometimes ago. IMHO it should really be in 2.6.27.
Patches 2 and 3 are very simple patches that IMHO may be merged in 2.6.27 without too much risks.
For the following patches, it may be better to wait for 2.6.28 as they are a little more complex. The last patch of the series also fixes something important, that is the detection of PCI busses (and hence PCI devices) on the BCM47xx SOC.
Signature · 4 lines
.''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
PATCH 1/6 BCM47xx: Add platform specific PCI code
This patch, ported from OpenWRT SVN, defines pcibios_map_irq() and pcibios_plat_dev_init() for the BCM47xx platform.
It fixes the regression introduced by commitaab547ce0d1493d400b6468c521a0137cd8c1edf.
--- arch/mips/pci/Makefile | 1 + arch/mips/pci/pci-bcm47xx.c | 58 +++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 59 insertions(+), 0 deletions(-) create mode 100644 arch/mips/pci/pci-bcm47xx.c
arch/mips/pci/Makefile +1 -0
@@ -15,6 +15,7 @@ obj-$(CONFIG_SOC_TX3927) += ops-tx3927.o obj-$(CONFIG_PCI_VR41XX) += ops-vr41xx.o pci-vr41xx.o obj-$(CONFIG_MARKEINS) += ops-emma2rh.o pci-emma2rh.o fixup-emma2rh.o obj-$(CONFIG_PCI_TX4927) += ops-tx4927.o +obj-$(CONFIG_BCM47XX) += pci-bcm47xx.o # # These are still pretty much in the old state, watch, go blind.
arch/mips/pci/pci-bcm47xx.c new file +58 -0
58 lines not on this page · raw message
Signature · 8 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
Hello.
Quoted 9 lines · Aurelien Jarno
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf.Signed-off-by Aurelien Jarno <[email protected]>Reviewed-by Michael Buesch <[email protected]>
[...]
Quoted 38 lines
arch/mips/pci/pci-bcm47xx.c new file +31
-0@@ -0,0 +1,58 @@ +/* +* Copyright (C) 2008 Michael Buesch <[email protected]> +* Copyright (C) 2008 Aurelien Jarno <[email protected]> +* +* This program is free software; you can redistribute it and/or modify it +* under the terms of the GNU General Public License as published by the +* Free Software Foundation; either version 2 of the License, or (at your +* option) any later version. +* +* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED +* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF +* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN +* NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, +* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT +* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF +* USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON +* ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF +* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +* +* You should have received a copy of the GNU General Public License along +* with this program; if not, write to the Free Software Foundation, Inc., +* 675 Mass Ave, Cambridge, MA 02139, USA. +*/ + +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/ssb/ssb.h> + +int pcibios_map_irq (const struct pci_dev *dev,u8 slot,u8 pin) +{
arch/mips/pci/pci-bcm47xx.c line 41
Quoted 10 lines
arch/mips/pci/pci-bcm47xx.c new file +10
-0+int res; + + res =ssb_pcibios_map_irq (dev, slot, pin); +if (res <0 ) { +printk (KERN_ALERT "PCI: Failed to map IRQ of device %s\n" , + dev->dev.bus_id); +return 0 ; + } +/* IRQ-0 and IRQ-1 are software interrupts. */ +WARN_ON ((res ==0 ) || (res ==1 ));
Unneeded ()...
Quoted 4 lines
arch/mips/pci/pci-bcm47xx.c new file +3
-0+ +return res; +}
arch/mips/pci/pci-bcm47xx.c line 54
Quoted 10 lines
arch/mips/pci/pci-bcm47xx.c new file +10
-0+ +int pcibios_plat_dev_init (struct pci_dev *dev) +{ +int err; + + err =ssb_pcibios_plat_dev_init (dev); +if (err) { +printk (KERN_ALERT "PCI: Failed to init device %s\n" , +pci_name (dev)); + }
Unneeded {}...
WBR, Sergei
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
Quoted 4 lines · On Thursday 25 September 2008 00:14:22 Sergei Shtylyov wrote:
arch/mips/pci/pci-bcm47xx.c line 41
Quoted 1 line
arch/mips/pci/pci-bcm47xx.c new file +1
-0+WARN_ON ((res ==0 ) || (res ==1 ));Unneeded ()...
Quoted 8 lines
arch/mips/pci/pci-bcm47xx.c line 54
Quoted 6 lines
arch/mips/pci/pci-bcm47xx.c new file +5
-0+ err =ssb_pcibios_plat_dev_init (dev); +if (err) { +printk (KERN_ALERT "PCI: Failed to init device %s\n" , +pci_name (dev)); + }Unneeded {}...
You must me kidding.
Get this freaking regression fix upstream, please. We're not going to respin it to fix trivial taste issues. (The above code is _perfectly_ fine and doesn't violate any rule).
Signature · 1 line
Greetings Michael.
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
In reply to Sergei ShtylyovQuoted 7 lines · Sergei Shtylyov
Quoted 5 lines · Aurelien Jarno
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf.
Did you read this part?
Quoted 5 lines
arch/mips/pci/pci-bcm47xx.c line 41
Quoted 3 lines
arch/mips/pci/pci-bcm47xx.c new file +2
-0+/* IRQ-0 and IRQ-1 are software interrupts. */ +WARN_ON ((res ==0 ) || (res ==1 ));Unneeded ()...
Pffttt...
Quoted 7 lines
arch/mips/pci/pci-bcm47xx.c line 54
Quoted 5 lines
arch/mips/pci/pci-bcm47xx.c new file +4
-0+if (err) { +printk (KERN_ALERT "PCI: Failed to init device %s\n" , +pci_name (dev)); + }Unneeded {}...
This is petty bullshit...
Is there a real reason why this hasn't been merged?
John
Signature · 2 lines
John W. Linville Linux should be at the core [email protected] of your literate lifestyle.
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
Hello.
Quoted 9 lines · John W. Linville
Quoted 6 lines
Quoted 6 lines
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf.Did you read this part?
I usually comment on the patches regardless of such stuff. Maybe I
shoudn't have this time in order to not take the blame for the patch bot being merged...
Quoted 11 lines
Quoted 8 lines
arch/mips/pci/pci-bcm47xx.c line 54
Quoted 6 lines
arch/mips/pci/pci-bcm47xx.c new file +4
-0+if (err) { +printk (KERN_ALERT "PCI: Failed to init device %s\n" , +pci_name (dev)); + }Unneeded {}...
This is petty bullshit...
No, just style nitpicking. :-)
Quoted 2 lines
Is there a real reason why this hasn't been merged?
This question is not to me.
Quoted 2 lines
John
WBR, Sergei
PATCH 1/6 BCM47xx: Add platform specific PCI code
In reply to Sergei ShtylyovFrom: Aurelien Jarno <[email protected]>
This patch, ported from OpenWRT SVN, defines pcibios_map_irq() and pcibios_plat_dev_init() for the BCM47xx platform.
It fixes the regression introduced by commit aab547ce0d1493d400b6468c521a0137cd8c1edf ("ssb: Add Gigabit Ethernet driver").
--- arch/mips/pci/Makefile | 1 + arch/mips/pci/pci-bcm47xx.c | 58 +++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 59 insertions(+), 0 deletions(-) create mode 100644 arch/mips/pci/pci-bcm47xx.c
arch/mips/pci/Makefile +1 -0
@@ -15,6 +15,7 @@ obj-$(CONFIG_SOC_TX3927) += ops-tx3927.o obj-$(CONFIG_PCI_VR41XX) += ops-vr41xx.o pci-vr41xx.o obj-$(CONFIG_MARKEINS) += ops-emma2rh.o pci-emma2rh.o fixup-emma2rh.o obj-$(CONFIG_PCI_TX4927) += ops-tx4927.o +obj-$(CONFIG_BCM47XX) += pci-bcm47xx.o # # These are still pretty much in the old state, watch, go blind.
arch/mips/pci/pci-bcm47xx.c new file +57 -0
57 lines not on this page · raw message
Signature · 8 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
Hello.
Quoted 5 lines · John W. Linville
From: Aurelien Jarno <[email protected]>
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.
Quoted 8 lines
It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf("ssb: Add Gigabit Ethernet driver").Signed-off-by Aurelien Jarno <[email protected]>Reviewed-by Michael Buesch <[email protected]>Signed-off-by John W. Linville <[email protected]>
Thanks for addessing my nitpicking (which you could have ignored :-).
I'm only seeing this patch posted on April 21st before yesterday (and
receiving no comments) -- probably Aurilen hasn't been persevering enough for Ralf to merge it. I've been a victim of Ralf's "forgetfullness" as well, so I can sympathise. :-)
WBR, Sergei
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
Quoted 18 lines · Sergei Shtylyov
Hello.
Quoted 5 lines · John W. Linville
From: Aurelien Jarno <[email protected]>
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.Quoted 8 lines
It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf("ssb: Add Gigabit Ethernet driver").Signed-off-by Aurelien Jarno <[email protected]>Reviewed-by Michael Buesch <[email protected]>Signed-off-by John W. Linville <[email protected]>
Quoted 5 lines
Thanks for addessing my nitpicking (which you could have ignored :-). I'm only seeing this patch posted on April 21st before yesterday (and
receiving no comments) -- probably Aurilen hasn't been persevering enough for Ralf to merge it. I've been a victim of Ralf's "forgetfullness" as well, so I can sympathise. :-)
You probably forget messages like:
http://www.linux-mips.org/archives/linux-mips/2008-05/msg00300.html
http://www.linux-mips.org/archives/linux-mips/2008-06/msg00128.html
and the numerous IRC pings.
Signature · 4 lines
.''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
Hello.
Quoted 11 lines · Aurelien Jarno
Quoted 10 lines
Quoted 10 lines
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf("ssb: Add Gigabit Ethernet driver").Signed-off-by Aurelien Jarno <[email protected]>Reviewed-by Michael Buesch <[email protected]>Signed-off-by John W. Linville <[email protected]>
Quoted 7 lines
Quoted 3 lines
Thanks for addessing my nitpicking (which you could have ignored :-).
I'm only seeing this patch posted on April 21st before yesterday (and
receiving no comments) -- probably Aurilen hasn't been persevering enough for Ralf to merge it. I've been a victim of Ralf's "forgetfullness" as well, so I can sympathise. :-)You probably forget messages like:
http://www.linux-mips.org/archives/linux-mips/2008-05/msg00300.html
http://www.linux-mips.org/archives/linux-mips/2008-06/msg00128.html
Yes, I have overlooked these since they had different subject.
Quoted 1 line
and the numerous IRC pings.
I've been thru that before. :-)
WBR, Sergei
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
Quoted 25 lines · On Thursday 25 September 2008 16:39:42 Sergei Shtylyov wrote:
Hello.
Quoted 17 lines · Aurelien Jarno
Quoted 11 lines
Quoted 10 lines
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf("ssb: Add Gigabit Ethernet driver").Signed-off-by Aurelien Jarno <[email protected]>Reviewed-by Michael Buesch <[email protected]>Signed-off-by John W. Linville <[email protected]>Quoted 5 lines
Thanks for addessing my nitpicking (which you could have ignored :-). I'm only seeing this patch posted on April 21st before yesterday (and
receiving no comments) -- probably Aurilen hasn't been persevering enough for Ralf to merge it. I've been a victim of Ralf's "forgetfullness" as well, so I can sympathise. :-)Quoted 3 lines
You probably forget messages like:
http://www.linux-mips.org/archives/linux-mips/2008-05/msg00300.html
http://www.linux-mips.org/archives/linux-mips/2008-06/msg00128.html
Quoted 5 lines
Yes, I have overlooked these since they had different subject.
Quoted 1 line
and the numerous IRC pings.
I've been thru that before. :-)
And one private discussion and the bugzilla entry. ;)
That's why I'm going to submit it to akpm or John, is it's not applied to the mips tree until October the 5th.
Signature · 1 line
Greetings Michael.
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
In reply to Sergei ShtylyovQuoted 18 lines · Sergei Shtylyov
Hello.
Quoted 5 lines · John W. Linville
From: Aurelien Jarno <[email protected]>
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.Quoted 8 lines
It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf("ssb: Add Gigabit Ethernet driver").Signed-off-by Aurelien Jarno <[email protected]>Reviewed-by Michael Buesch <[email protected]>Signed-off-by John W. Linville <[email protected]>
Quoted 5 lines
Thanks for addessing my nitpicking (which you could have ignored :-). I'm only seeing this patch posted on April 21st before yesterday (and
receiving no comments) -- probably Aurilen hasn't been persevering enough for Ralf to merge it. I've been a victim of Ralf's "forgetfullness" as well, so I can sympathise. :-)
Sergei, I'm sorry if I was rude to you.
Ralf, given the OpenWRT and SSB connection, I could probably merge this series through my tree if you don't want to take it in yours for some reason.
John
Signature · 2 lines
John W. Linville Linux should be at the core [email protected] of your literate lifestyle.
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
Quoted 25 lines · On Thursday 25 September 2008 19:24:30 John W. Linville wrote:
Quoted 18 lines · Sergei Shtylyov
Hello.
Quoted 5 lines · John W. Linville
From: Aurelien Jarno <[email protected]>
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.Quoted 8 lines
It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf("ssb: Add Gigabit Ethernet driver").Signed-off-by Aurelien Jarno <[email protected]>Reviewed-by Michael Buesch <[email protected]>Signed-off-by John W. Linville <[email protected]>Quoted 5 lines
Thanks for addessing my nitpicking (which you could have ignored :-). I'm only seeing this patch posted on April 21st before yesterday (and
receiving no comments) -- probably Aurilen hasn't been persevering enough for Ralf to merge it. I've been a victim of Ralf's "forgetfullness" as well, so I can sympathise. :-)
Quoted 5 lines
Sergei, I'm sorry if I was rude to you.
Ralf, given the OpenWRT and SSB connection, I could probably merge this series through my tree if you don't want to take it in yours for some reason.
Yeah, the regression also made it through your tree, so it's probably OK to merge the fix through it, too. Let's give 'em another week and merge it through your tree if it's still not in the mips tree.
Signature · 1 line
Greetings Michael.
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
In reply to Sergei ShtylyovQuoted 11 lines · Sergei Shtylyov
arch/mips/pci/pci-bcm47xx.c line 41
Quoted 9 lines
arch/mips/pci/pci-bcm47xx.c new file +8
-0+ res =ssb_pcibios_map_irq (dev, slot, pin); +if (res <0 ) { +printk (KERN_ALERT "PCI: Failed to map IRQ of device %s\n" , + dev->dev.bus_id); +return 0 ; + } +/* IRQ-0 and IRQ-1 are software interrupts. */ +WARN_ON ((res ==0 ) || (res ==1 ));Unneeded ()...
My rule of thumb for the use of parenthesis is that the reader of a piece of code should not have to know much about operator precendence, so the occasional avoidable parenthesis if it serves readability.
Ralf
Re: [PATCH 1/6] [MIPS] BCM47xx: Add platform specific PCI code
In reply to Aurelien JarnoQuoted 8 lines · On Wednesday 24 September 2008 21:19:55 Aurelien Jarno wrote:
This patch, ported from OpenWRT SVN, defines
pcibios_map_irq()andpcibios_plat_dev_init()for the BCM47xx platform.It fixes the regression introduced by commit
aab547ce0d1493d400b6468c521a0137cd8c1edf.Signed-off-by Aurelien Jarno <[email protected]>Reviewed-by Michael Buesch <[email protected]>
Is this STILL not merged?
Come on mips guys, WAKE UP!
This compiletime-regression is fixed since several months and this is the third time (at least) that we submit this patch.
A kernel.org bugzilla entry does also exist.
I'll bypass the mips tree and send it to akpm, if this takes any further delay.
Quoted 86 lines
--- arch/mips/pci/Makefile | 1 + arch/mips/pci/pci-bcm47xx.c | 58 +++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 59 insertions(+), 0 deletions(-) create mode 100644 arch/mips/pci/pci-bcm47xx.carch/mips/pci/Makefile +1
-0@@ -15,6 +15,7 @@ obj-$(CONFIG_SOC_TX3927) += ops-tx3927.o obj-$(CONFIG_PCI_VR41XX) += ops-vr41xx.o pci-vr41xx.o obj-$(CONFIG_MARKEINS) += ops-emma2rh.o pci-emma2rh.o fixup-emma2rh.o obj-$(CONFIG_PCI_TX4927) += ops-tx4927.o +obj-$(CONFIG_BCM47XX) += pci-bcm47xx.o # # These are still pretty much in the old state, watch, go blind.arch/mips/pci/pci-bcm47xx.c new file +58
-0@@ -0,0 +1,58 @@ +/* +* Copyright (C) 2008 Michael Buesch <[email protected]> +* Copyright (C) 2008 Aurelien Jarno <[email protected]> +* +* This program is free software; you can redistribute it and/or modify it +* under the terms of the GNU General Public License as published by the +* Free Software Foundation; either version 2 of the License, or (at your +* option) any later version. +* +* THIS SOFTWARE IS PROVIDED ``AS IS'' AND ANY EXPRESS OR IMPLIED +* WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF +* MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN +* NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, +* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT +* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF +* USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON +* ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF +* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. +* +* You should have received a copy of the GNU General Public License along +* with this program; if not, write to the Free Software Foundation, Inc., +* 675 Mass Ave, Cambridge, MA 02139, USA. +*/ + +#include <linux/types.h> +#include <linux/pci.h> +#include <linux/ssb/ssb.h> + +int pcibios_map_irq (const struct pci_dev *dev,u8 slot,u8 pin) +{ +int res; + + res =ssb_pcibios_map_irq (dev, slot, pin); +if (res <0 ) { +printk (KERN_ALERT "PCI: Failed to map IRQ of device %s\n" , + dev->dev.bus_id); +return 0 ; + } +/* IRQ-0 and IRQ-1 are software interrupts. */ +WARN_ON ((res ==0 ) || (res ==1 )); + +return res; +} + +int pcibios_plat_dev_init (struct pci_dev *dev) +{ +int err; + + err =ssb_pcibios_plat_dev_init (dev); +if (err) { +printk (KERN_ALERT "PCI: Failed to init device %s\n" , +pci_name (dev)); + } + +return err; +} +Signature · 1 line
1.5.6.5
Signature · 1 line
Greetings Michael.
PATCH v2 1/6 BCM47xx: Add platform specific PCI code
In reply to Aurelien JarnoPlease find below the version 2 of this patch that addresses all he comments. It removes the unneeded braces and brackets, and add __init to pcibios_map_irq().
---
This patch, ported from OpenWRT SVN, defines pcibios_map_irq() and pcibios_plat_dev_init() for the BCM47xx platform.
It fixes the regression introduced by commitaab547ce0d1493d400b6468c521a0137cd8c1edf.
--- arch/mips/pci/Makefile | 1 + arch/mips/pci/pci-bcm47xx.c | 57 +++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 58 insertions(+), 0 deletions(-) create mode 100644 arch/mips/pci/pci-bcm47xx.c
arch/mips/pci/Makefile +1 -0
@@ -15,6 +15,7 @@ obj-$(CONFIG_SOC_TX3927) += ops-tx3927.o obj-$(CONFIG_PCI_VR41XX) += ops-vr41xx.o pci-vr41xx.o obj-$(CONFIG_MARKEINS) += ops-emma2rh.o pci-emma2rh.o fixup-emma2rh.o obj-$(CONFIG_PCI_TX4927) += ops-tx4927.o +obj-$(CONFIG_BCM47XX) += pci-bcm47xx.o # # These are still pretty much in the old state, watch, go blind.
arch/mips/pci/pci-bcm47xx.c new file +57 -0
57 lines not on this page · raw message
Signature · 8 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
PATCH v3 1/6 BCM47xx: Add platform specific PCI code
Please find below the version 3 of this patch that addresses all he comments. This new version moves all the code to pcibios_plat_dev_init() from pcibios_map_irq() to support dynamic interrupt routing. This way errors can be reported correctly.
---
This patch defines pcibios_map_irq() and pcibios_plat_dev_init() for the BCM47xx platform.
It fixes the regression introduced by commitaab547ce0d1493d400b6468c521a0137cd8c1edf.
--- arch/mips/pci/Makefile | 1 + arch/mips/pci/pci-bcm47xx.c | 60 +++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 61 insertions(+), 0 deletions(-) create mode 100644 arch/mips/pci/pci-bcm47xx.c
arch/mips/pci/Makefile +1 -0
@@ -15,6 +15,7 @@ obj-$(CONFIG_SOC_TX3927) += ops-tx3927.o obj-$(CONFIG_PCI_VR41XX) += ops-vr41xx.o pci-vr41xx.o obj-$(CONFIG_MARKEINS) += ops-emma2rh.o pci-emma2rh.o fixup-emma2rh.o obj-$(CONFIG_PCI_TX4927) += ops-tx4927.o +obj-$(CONFIG_BCM47XX) += pci-bcm47xx.o # # These are still pretty much in the old state, watch, go blind.
arch/mips/pci/pci-bcm47xx.c new file +60 -0
60 lines not on this page · raw message
Signature · 7 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
Re: [PATCH 1/6 v3] [MIPS] BCM47xx: Add platform specific PCI code
Quoted 4 lines · Aurelien Jarno
Please find below the version 3 of this patch that addresses all he comments. This new version moves all the code to
pcibios_plat_dev_init()frompcibios_map_irq()to support dynamic interrupt routing. This way errors can be reported correctly.
This one looks good. Thanks, applied.
Ralf
PATCH 2/6 WGT634U: Add machine detection message
In reply to Aurelien JarnoThis adds a printk message when a WGT634U machine is detected.
--- arch/mips/bcm47xx/wgt634u.c | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-)
arch/mips/bcm47xx/wgt634u.c +3 -0
@@ -112,6 +112,9 @@ static int __init wgt634u_init(void) ((et0mac[1] == 0x09 && et0mac[2] == 0x5b) || (et0mac[1] == 0x0f && et0mac[2] == 0xb5))) {struct ssb_mipscore *mcore = &ssb_bcm47xx.mipscore; + +printk (KERN_INFO "WGT634U machine detected.\n" ); + wgt634u_flash_data.width = mcore->flash_buswidth; wgt634u_flash_resource.start = mcore->flash_window; wgt634u_flash_resource.end = mcore->flash_window
Signature · 7 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
PATCH 3/6 Remove references to BCM947XX
In reply to Aurelien JarnoThis patch removes the remaining reference
to the BCM947xx development board codename.
--- include/asm-mips/mach-bcm47xx/war.h | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-)
include/asm-mips/mach-bcm47xx/war.h +3 -3
@@ -5,8 +5,8 @@* * Copyright (C) 2002, 2004, 2007 by Ralf Baechle <[email protected]> */ -#ifndef __ASM_MIPS_MACH_BCM947XX_WAR_H -#define __ASM_MIPS_MACH_BCM947XX_WAR_H +#ifndef __ASM_MIPS_MACH_BCM47XX_WAR_H +#define __ASM_MIPS_MACH_BCM47XX_WAR_H#define R4600_V1_INDEX_ICACHEOP_WAR 0#define R4600_V1_HIT_CACHEOP_WAR 0
@@ -22,4 +22,4 @@#define R10000_LLSC_WAR 0#define MIPS34K_MISSED_ITLB_WAR 0 -#endif /* __ASM_MIPS_MACH_BCM947XX_WAR_H */ +#endif /* __ASM_MIPS_MACH_BCM47XX_WAR_H */
Signature · 8 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
PATCH 4/6 BCM47xx: Use the new SSB GPIO API
In reply to Aurelien JarnoThis patch simplifies the BCM47xx GPIO code by using the new SSB GPIO API, which does a lot things that were implemented directly in the BCM47xx code.
--- arch/mips/Kconfig | 1 + arch/mips/bcm47xx/gpio.c | 85 +++++++++++++-------------------- arch/mips/bcm47xx/setup.c | 5 +- include/asm-mips/mach-bcm47xx/gpio.h | 41 ++++++++-------- 4 files changed, 58 insertions(+), 74 deletions(-)
arch/mips/Kconfig +1 -0
@@ -52,6 +52,7 @@ config BCM47XX select SSB select SSB_DRIVER_MIPS select SSB_DRIVER_EXTIF + select SSB_EMBEDDED select SSB_PCICORE_HOSTMODE if PCI select GENERIC_GPIO select SYS_HAS_EARLY_PRINTK
arch/mips/bcm47xx/gpio.c +34 -51
102 lines not on this page · raw message
arch/mips/bcm47xx/setup.c +3 -2
@@ -27,6 +27,7 @@#include <linux/types.h> #include <linux/ssb/ssb.h> +#include <linux/ssb/ssb_embedded.h> #include <asm/bootinfo.h> #include <asm/reboot.h> #include <asm/time.h>
@@ -41,7 +42,7 @@ static void bcm47xx_machine_restart(char *command)printk (KERN_ALERT "Please stand by while rebooting the system...\n" );local_irq_disable ();/* Set the watchdog timer to reset immediately */ -ssb_chipco_watchdog_timer_set (&ssb_bcm47xx.chipco,1 ); +ssb_watchdog_timer_set (&ssb_bcm47xx,1 );while (1 )cpu_relax (); }
@@ -50,7 +51,7 @@ static void bcm47xx_machine_halt(void) {/* Disable interrupts and watchdog and spin forever */ local_irq_disable (); -ssb_chipco_watchdog_timer_set (&ssb_bcm47xx.chipco,0 ); +ssb_watchdog_timer_set (&ssb_bcm47xx,0 );while (1 )cpu_relax (); }
include/asm-mips/mach-bcm47xx/gpio.h +20 -21
67 lines not on this page · raw message
Signature · 8 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
PATCH 5/6 Add WGT634U reset button support
In reply to Aurelien JarnoThis patch adds support for the reset button of WGT634U machine, using GPIO interrupts. Based on a patch from Michel Lespinasse.
--- arch/mips/bcm47xx/wgt634u.c | 37 +++++++++++++++++++++++++++++++++++++ 1 files changed, 37 insertions(+), 0 deletions(-)
arch/mips/bcm47xx/wgt634u.c +37 -0
@@ -11,6 +11,9 @@#include <linux/leds.h> #include <linux/mtd/physmap.h> #include <linux/ssb/ssb.h> +#include <linux/interrupt.h> +#include <linux/reboot.h> +#include <asm/gpio.h> #include <asm/mach-bcm47xx/bcm47xx.h> /* GPIO definitions for the WGT634U */
@@ -99,6 +102,30 @@ static struct platform_device *wgt634u_devices[] __initdata = { &wgt634u_gpio_leds, }; +static irqreturn_t gpio_interrupt (int irq,void *ignored) +{ +int state; + +/* Interrupts are shared, check if the current one is +a GPIO interrupt. */ +if (!ssb_chipco_irq_status (&ssb_bcm47xx.chipco, +SSB_CHIPCO_IRQ_GPIO )) +return IRQ_NONE ; + + state =gpio_get_value (WGT634U_GPIO_RESET ); + +/* Interrupt are level triggered, revert the interrupt polarity +to clear the interrupt. */ +gpio_polarity (WGT634U_GPIO_RESET , state); + +if (!state) { +printk (KERN_INFO "Reset button pressed" ); +ctrl_alt_del (); + } + +return IRQ_HANDLED ; +} +static int __init wgt634u_init (void ) {/* There is no easy way to detect that we are running on a WGT634U
@@ -115,6 +142,16 @@ static int __init wgt634u_init(void)printk (KERN_INFO "WGT634U machine detected.\n" ); +if (!request_irq (gpio_to_irq (WGT634U_GPIO_RESET ), + gpio_interrupt,IRQF_SHARED , +"WGT634U GPIO" , &ssb_bcm47xx.chipco)) { +gpio_direction_input (WGT634U_GPIO_RESET ); +gpio_intmask (WGT634U_GPIO_RESET ,1 ); +ssb_chipco_irq_mask (&ssb_bcm47xx.chipco, +SSB_CHIPCO_IRQ_GPIO , +SSB_CHIPCO_IRQ_GPIO ); + } + wgt634u_flash_data.width = mcore->flash_buswidth; wgt634u_flash_resource.start = mcore->flash_window; wgt634u_flash_resource.end = mcore->flash_window
Signature · 8 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net
PATCH 6/6 Scan PCI busses when they are registered
In reply to Aurelien JarnoThe patch below changes register_pci_controller() such that controllers being added after pcibios_init() has run are be scanned immediately.
This is needed for example by the BCM47xx PCI controller, which is located on the SSB bus, which is now initialized after the PCI subsystem.
--- arch/mips/pci/pci.c | 80 +++++++++++++++++++++++++++++++++----------------- 1 files changed, 53 insertions(+), 27 deletions(-)
arch/mips/pci/pci.c +53 -27
@@ -34,6 +34,8 @@ static struct pci_controller *hose_head, **hose_tail = &hose_head;unsigned long PCIBIOS_MIN_IO =0x0000 ;unsigned long PCIBIOS_MIN_MEM =0 ; +static int pci_initialized; +/* * We need to avoid collisions with `mirrored' VGA ports * and other strange ISA hardware, so we always want the
@@ -74,6 +76,42 @@ pcibios_align_resource(void *data, struct resource *res, res->start = start; } +static void __devinit pcibios_scanbus (struct pci_controller *hose) +{ +static int next_busno; +static int need_domain_info; +struct pci_bus *bus; + +if (!hose->iommu) +PCI_DMA_BUS_IS_PHYS =1 ; + +if (hose->get_busno && pci_probe_only) + next_busno = (*hose->get_busno)(); + + bus =pci_scan_bus (next_busno, hose->pci_ops, hose); + hose->bus = bus; + + need_domain_info = need_domain_info || hose->index; + hose->need_domain_info = need_domain_info; +if (bus) { + next_busno = bus->subordinate +1 ; +/* Don't allow 8-bit bus number overflow inside the hose - +reserve some space for bridges. */ +if (next_busno >224 ) { + next_busno =0 ; + need_domain_info =1 ; + } + +if (!pci_probe_only) { +pci_bus_size_bridges (bus); +pci_bus_assign_resources (bus); +pci_enable_bridges (bus); + } + } +} + +static DEFINE_MUTEX (pci_scan_mutex); +void __devinit register_pci_controller (struct pci_controller *hose) {if (request_resource (&iomem_resource, hose->mem_resource) <0 )
@@ -93,6 +131,17 @@ void __devinit register_pci_controller(struct pci_controller *hose)printk (KERN_WARNING "registering PCI controller with io_map_base unset\n" ); } + +/* +* Scan the bus if it is register after the PCI subsystem +* initialization. +*/ +if (pci_initialized) { +mutex_lock (&pci_scan_mutex); +pcibios_scanbus (hose); +mutex_unlock (&pci_scan_mutex); + } +return ; out:
@@ -125,38 +174,15 @@ static u8 __init common_swizzle(struct pci_dev *dev, u8 *pinp)static int __init pcibios_init (void ) {struct pci_controller *hose; -struct pci_bus *bus; -int next_busno; -int need_domain_info =0 ;/* Scan all of the recorded PCI controllers. */ -for (next_busno =0 , hose = hose_head; hose; hose = hose->next) { - -if (!hose->iommu) -PCI_DMA_BUS_IS_PHYS =1 ; - -if (hose->get_busno && pci_probe_only) - next_busno = (*hose->get_busno)(); - - bus =pci_scan_bus (next_busno, hose->pci_ops, hose); - hose->bus = bus; - need_domain_info = need_domain_info || hose->index; - hose->need_domain_info = need_domain_info; -if (bus) { - next_busno = bus->subordinate +1 ; -/* Don't allow 8-bit bus number overflow inside the hose - -reserve some space for bridges. */ -if (next_busno >224 ) { - next_busno =0 ; - need_domain_info =1 ; - } - } - } +for (hose = hose_head; hose; hose = hose->next) +pcibios_scanbus (hose); -if (!pci_probe_only) -pci_assign_unassigned_resources ();pci_fixup_irqs (common_swizzle, pcibios_map_irq); + pci_initialized =1 ; +return 0 ; }
Signature · 8 lines
1.5.6.5 -- .''`. Aurelien Jarno | GPG: 1024D/F1BCDB73 : :' : Debian developer | Electrical Engineer `. `' [email protected] | [email protected] `- people.debian.org/~aurel32 | www.aurel32.net