Skip to contentSkip to conversation

x86/pci: Broken build for X86_VISWS

PATCH New Robert Richter 2 participants 7 messages Jul 10, 2008 x86

lore
Share & tools
lore
#0 lore raw

Ingo,

during testing of my pci changes I observed a build error if X86_VISWS is set:

arch/x86/mach-visws/built-in.o: In function `machine_power_off':
(.text+0x48): undefined reference to `pci_bus0' make[1]: *** [.tmp_vmlinux1] Error 1
make: *** [sub-make] Error 2

The root cause is disabling PCI dependencies in change 1ac97018 that disables compilation of x86/pci for VISWS though there are files in needed to build. I am not sure if it is save to enable PCI for VISWS to fix this.

Could you take a look at this?

Thanks,

-Robert

Signature · 3 lines
Advanced Micro Devices, Inc.
Operating System Research Center
email: [email protected]
#1 lore raw

PATCH 1/1 x86/pci: Fix build error for X86_VISWS

---
 arch/x86/Kconfig |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
arch/x86/Kconfig +1 -1
index bb0c0d0..51329ee 100644
@@ -283,7 +283,7 @@ config X86_VOYAGER
 
 config X86_VISWS
 	bool "SGI 320/540 (Visual Workstation)"
-	depends on X86_32 && !PCI
+	depends on X86_32 && PCI
 	help
 	  The SGI Visual Workstation series is an IA32-based workstation
 	  based on SGI systems chips with some legacy PC hardware attached.
Signature · 1 line
1.5.5.4
#2 lore raw
In reply to Robert Richter
Quoted 10 lines · Robert Richter

Ingo,

during testing of my pci changes I observed a build error if X86_VISWS is set:

arch/x86/mach-visws/built-in.o: In function `machine_power_off':
(.text+0x48): undefined reference to `pci_bus0' make[1]: *** [.tmp_vmlinux1] Error 1
make: *** [sub-make] Error 2

Quoted 6 lines

The root cause is disabling PCI dependencies in change 1ac97018 that disables compilation of x86/pci for VISWS though there are files in needed to build. I am not sure if it is save to enable PCI for VISWS to fix this.

Could you take a look at this?

well spotted. Does the patch below fix it for you?

Ingo

------------------------->
commit 50e810529abfb39259c2877ed8dc784a4bb81167

Author: Ingo Molnar <[email protected]>
Date:   Thu Jul 10 12:21:58 2008 +0200

x86: fix visws and vsmp build

these two sub-architectures want PCI to be default-on, not default-off.

---
 arch/x86/Kconfig |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)
arch/x86/Kconfig +2 -2
index 4d85501..c8f7fb6 100644
@@ -287,7 +287,7 @@ config X86_VOYAGER
 
 config X86_VISWS
 	bool "SGI 320/540 (Visual Workstation)"
-	depends on X86_32 && !PCI
+	depends on X86_32 && PCI
 	help
 	  The SGI Visual Workstation series is an IA32-based workstation
 	  based on SGI systems chips with some legacy PC hardware attached.
@@ -359,7 +359,7 @@ config X86_RDC321X
 config X86_VSMP
 	bool "Support for ScaleMP vSMP"
 	select PARAVIRT
-	depends on X86_64 && !PCI
+	depends on X86_64 && PCI
 	help
 	  Support for ScaleMP vSMP systems.  Say 'Y' here if this kernel is
 	  supposed to run on these EM64T-based machines.  Only choose this option
#3 lore raw
Quoted 12 lines · Ingo Molnar
Quoted 10 lines · Robert Richter

Ingo,

during testing of my pci changes I observed a build error if X86_VISWS is set:

arch/x86/mach-visws/built-in.o: In function `machine_power_off':
(.text+0x48): undefined reference to `pci_bus0' make[1]: *** [.tmp_vmlinux1] Error 1
make: *** [sub-make] Error 2

Quoted 8 lines
Quoted 6 lines

The root cause is disabling PCI dependencies in change 1ac97018 that disables compilation of x86/pci for VISWS though there are files in needed to build. I am not sure if it is save to enable PCI for VISWS to fix this.

Could you take a look at this?

well spotted. Does the patch below fix it for you?

there's another visws build bug related to generic-ipi - i fixed that in tip/master.

Ingo

#4 lore raw
In reply to Ingo Molnar
Quoted 1 line · Ingo Molnar

well spotted. Does the patch below fix it for you?

Yes, it should. I send out a fix with the same change and it builds well. But I did not realize that X86_VSMP was also affected.

-Robert

Signature · 3 lines
Advanced Micro Devices, Inc.
Operating System Research Center
email: [email protected]
#5 lore raw
Quoted 5 lines · Robert Richter
Quoted 1 line · Ingo Molnar

well spotted. Does the patch below fix it for you?

Yes, it should. I send out a fix with the same change and it builds well. But I did not realize that X86_VSMP was also affected.

i'm wondering why you werent hit by the generic-ipi build bug. Perhaps you had SMP disabled?

Ingo

#6 lore raw
Quoted 2 lines · Ingo Molnar

i'm wondering why you werent hit by the generic-ipi build bug. Perhaps you had SMP disabled?

No:

$ grep 'SMP\|VISWS' configs/linux-i386-visws.config 
CONFIG_X86_SMP=y
CONFIG_X86_32_SMP=y
CONFIG_SMP=y
CONFIG_X86_FIND_SMP_CONFIG=y
CONFIG_X86_VISWS=y
# CONFIG_X86_VSMP is not set
CONFIG_X86_VISWS_APIC=y

Probably it is because I am on the tip/x86/unify-pci tree where the IPI changes were not in.

-Robert

Signature · 3 lines
Advanced Micro Devices, Inc.
Operating System Research Center
email: [email protected]