Skip to contentSkip to conversation

x86: Fix regression build issue of coreboot-x86_defconfig

PATCH New Bin Meng 4 participants 16 messages May 27, 2015 – May 29, 2015 1 message not in the archive

lore
Share & tools
lore

Bin Meng

author
#0 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Signature · 1 line
1.8.2.1
author
#1 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Simon,

Quoted 15 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +0 -0
index 66f94d0..799853f 100644

configs/coreboot-x86_defconfig

Quoted 6 lines
configs/coreboot-x86_defconfig +1 -1
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

Regards,
Bin

#2 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

You need to remember to run savedefconfig when changing Kconfig or defconfig.

Cheers,
-Joe

author
#3 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Joe,

Quoted 35 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

Quoted 5 lines

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

Regards,
Bin

#4 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Bin,

Quoted 42 lines · Bin Meng

Hi Joe,

Quoted 38 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 7 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

Cheers,
-Joe

#5 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi,

Quoted 61 lines · Joe Hershberger

Hi Bin,

Quoted 42 lines · Bin Meng

Hi Joe,

Quoted 38 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 7 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Quoted 5 lines

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

Regards,
Simon

#6 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Simon,

Quoted 64 lines · Simon Glass

Hi,

Quoted 61 lines · Joe Hershberger

Hi Bin,

Quoted 42 lines · Bin Meng

Hi Joe,

Quoted 38 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 7 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Quoted 8 lines
Quoted 5 lines

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

It should also include removing the (now) redundant CONFIG_VENDOR_EMULATION from that defconfig. That is best done by running savedefconfig on that defconfig after applying the "x86: Make QEMU the default vendor" patch.

Bin, are you in the middle of this?

Cheers,
-Joe

author
#7 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Joe,

Quoted 67 lines · Joe Hershberger

Hi Simon,

Quoted 64 lines · Simon Glass

Hi,

Quoted 61 lines · Joe Hershberger

Hi Bin,

Quoted 42 lines · Bin Meng

Hi Joe,

Quoted 38 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 7 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Thanks for the explanation. Now it is clear to me.

Quoted 9 lines
Quoted 8 lines
Quoted 5 lines

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

Quoted 7 lines

It should also include removing the (now) redundant CONFIG_VENDOR_EMULATION from that defconfig. That is best done by running savedefconfig on that defconfig after applying the "x86: Make QEMU the default vendor" patch.

Bin, are you in the middle of this?

Yes, now we should remove CONFIG_VENDOR_EMULATION from configs/qemu-x86_defconfig. But if removing this, I feel this is inconsistent with other x86 boards' defconfig. Is this really a must-have to replace the defconfig with the one generated by 'savedefconfig'? If so, can we make it part of checkpatch.pl so that each time we submit the patch which contains modifications to the defconfig we can make sure the file is clean?

Regards,
Bin

author
#8 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

In reply to Simon Glass

Hi Simon,

Quoted 64 lines · Simon Glass

Hi,

Quoted 61 lines · Joe Hershberger

Hi Bin,

Quoted 42 lines · Bin Meng

Hi Joe,

Quoted 38 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 7 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Quoted 9 lines
Quoted 5 lines

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

It looks OK. Another thing, I noticed this patch [1] is not in the testing branch. Did you apply it to somewhere else?

[1] http://patchwork.ozlabs.org/patch/472988/

Regards,
Bin

#9 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

HI Bin,

Quoted 7 lines · Bin Meng

Hi Simon,

Quoted 3 lines · Simon Glass

Hi,

On 27 May 2015 at 10:27, Joe Hershberger <[email protected]>

Quoted 11 lines · wrote:
Quoted 11 lines
Quoted 11 lines

Hi Bin,

Quoted 7 lines · Bin Meng

Hi Joe,

Quoted 3 lines · Joe Hershberger

Hi Bin,

On Wed, May 27, 2015 at 11:01 AM, Bin Meng <[email protected]>

Quoted 3 lines · wrote:
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

Hi Simon,

On Wed, May 27, 2015 at 11:55 PM, Bin Meng <[email protected]>

Quoted 1 line · wrote:
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

Commit bd328eb "Clean all defconfigs with savedefconfig"

accidentally

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

removed 'CONFIG_VENDOR_COREBOOT=y' from

configs/coreboot-x86_defconfig.

Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines

Review

Quoted 9 lines

This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
(quoted diff) +0 -0

b/configs/coreboot-x86_defconfig

Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines

index 66f94d0..799853f 100644

configs/coreboot-x86_defconfig +0 -0
@@ -1,4 +1,5 @@
 CONFIG_X86=y
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

configs/coreboot-x86_defconfig

Quoted 5 lines
configs/coreboot-x86_defconfig +1 -1
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
--

Please apply this patch after commit

http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

and before commit

http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc
.

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line

I've verified the build with buildman on a new 'testing' branch

with

Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 1 line

insertion of this patch.

This should be squashed as part of

http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or

defconfig.

Quoted 20 lines
Quoted 20 lines
Quoted 19 lines
Quoted 6 lines
Quoted 1 line

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

Quoted 6 lines
Quoted 3 lines

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

It looks OK. Another thing, I noticed this patch [1] is not in the testing branch. Did you apply it to somewhere else?

I was worried that it was submitted after the merge window and affects other code.

Quoted 5 lines

[1] http://patchwork.ozlabs.org/patch/472988/

Regards,
Bin

Regards,
Simon

author
#10 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Simon,

Quoted 92 lines · Simon Glass

HI Bin,

Quoted 89 lines · Bin Meng

Hi Simon,

Quoted 84 lines · Simon Glass

Hi,

Quoted 71 lines · Joe Hershberger

Hi Bin,

Quoted 51 lines · Bin Meng

Hi Joe,

Quoted 47 lines · Joe Hershberger

Hi Bin,

Quoted 31 lines · Bin Meng

Hi Simon,

Quoted 18 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally
removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig.
This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +0 -0
index 66f94d0..799853f 100644

configs/coreboot-x86_defconfig

Quoted 6 lines
configs/coreboot-x86_defconfig +1 -1
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
--

Please apply this patch after commit

Quoted 7 lines

http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit

http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with
insertion of this patch.

This should be squashed as part of

http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 8 lines
Quoted 3 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Quoted 5 lines

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

Quoted 5 lines
Quoted 2 lines

It looks OK. Another thing, I noticed this patch [1] is not in the testing branch. Did you apply it to somewhere else?

I was worried that it was submitted after the merge window and affects other code.

OK, understood. But I think it is safe to remove given no board is using it.

Quoted 3 lines
Quoted 3 lines

[1] http://patchwork.ozlabs.org/patch/472988/

Regards,
Bin

author
#11 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

In reply to Simon Glass

Hi Simon, Hi Joe,

Quoted 64 lines · Simon Glass

Hi,

Quoted 61 lines · Joe Hershberger

Hi Bin,

Quoted 42 lines · Bin Meng

Hi Joe,

Quoted 38 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 7 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Quoted 9 lines
Quoted 5 lines

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

I just found u-boot.rom created from u-boot-x86/testing branch does not have working network any more on QEMU. It is working on the u-boot-x86/master branch.

=> set serverip 10.10.0.100;set ipaddr 10.10.0.108;set netmask 255.255.255.0;set gatewayip 10.10.0.100
=> run ramboot
*** ERROR: `serverip' not set
*** ERROR: `serverip' not set

But 'print' says the env indeed has the 'serverip'. I think there should be something in the upstream that caused this after the u-boot-x86/testing branch rebase. Do you know where might be the problem?

Regards,
Bin

#12 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Bin,

Quoted 10 lines

I just found u-boot.rom created from u-boot-x86/testing branch does not have working network any more on QEMU. It is working on the u-boot-x86/master branch.

=> set serverip 10.10.0.100;set ipaddr 10.10.0.108;set netmask 255.255.255.0;set gatewayip 10.10.0.100
=> run ramboot
*** ERROR: `serverip' not set
*** ERROR: `serverip' not set

Quoted 4 lines

But 'print' says the env indeed has the 'serverip'. I think there should be something in the upstream that caused this after the u-boot-x86/testing branch rebase. Do you know where might be the problem?

This looks very similar to an issue that got patched on the i.MX6 yesterday by Stefano which said:

commit fd3056337e6fcc140f400e11edd33f6f1cb37de1
"Use env callbacks for net variables" has a side effect on i.MX6 boards because they do not set CONFIG_NET: the ip address results not set, but it is stored in the environment."

Might be something to check...

Regards,

Andy.

#13 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

In reply to Bin Meng

Hi Bin,

Quoted 87 lines · Bin Meng

Hi Simon, Hi Joe,

Quoted 64 lines · Simon Glass

Hi,

Quoted 61 lines · Joe Hershberger

Hi Bin,

Quoted 42 lines · Bin Meng

Hi Joe,

Quoted 38 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 7 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Quoted 9 lines
Quoted 5 lines

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

I just found u-boot.rom created from u-boot-x86/testing branch does not have working network any more on QEMU. It is working on the u-boot-x86/master branch.

=> set serverip 10.10.0.100;set ipaddr 10.10.0.108;set netmask 255.255.255.0;set gatewayip 10.10.0.100
=> run ramboot
*** ERROR: `serverip' not set
*** ERROR: `serverip' not set

Quoted 4 lines

But 'print' says the env indeed has the 'serverip'. I think there should be something in the upstream that caused this after the u-boot-x86/testing branch rebase. Do you know where might be the problem?

I'll hold off pushing to master for now.

Regards,
Simon

author
#14 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Simon,

Quoted 90 lines · Simon Glass

Hi Bin,

Quoted 87 lines · Bin Meng

Hi Simon, Hi Joe,

Quoted 64 lines · Simon Glass

Hi,

Quoted 61 lines · Joe Hershberger

Hi Bin,

Quoted 42 lines · Bin Meng

Hi Joe,

Quoted 38 lines · Joe Hershberger

Hi Bin,

Quoted 23 lines · Bin Meng

Hi Simon,

Quoted 20 lines · Bin Meng

Commit bd328eb "Clean all defconfigs with savedefconfig" accidentally removed 'CONFIG_VENDOR_COREBOOT=y' from configs/coreboot-x86_defconfig. This commit reverts the change.

---

 configs/coreboot-x86_defconfig | 1 +
 1 file changed, 1 insertion(+)
configs/coreboot-x86_defconfig +1 -0
index 66f94d0..799853f 100644
@@ -1,4 +1,5 @@
 CONFIG_X86=y
+CONFIG_VENDOR_COREBOOT=y
 CONFIG_TARGET_COREBOOT=y
 CONFIG_OF_CONTROL=y
 CONFIG_DM_PCI=y
Quoted 8 lines
Quoted 1 line

--

Please apply this patch after commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=3506805839e14a67b0971b02c7784e37b85d5fbf
and before commit
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc.
I've verified the build with buildman on a new 'testing' branch with insertion of this patch.

This should be squashed as part of
http://git.denx.de/?p=u-boot/u-boot-x86.git;a=commit;h=05cab1d6da2a84911fa0ec0ffa8fa038adef4dbc

Quoted 7 lines
Quoted 2 lines

You need to remember to run savedefconfig when changing Kconfig or defconfig.

I still don't get it. commit 65c4ac0 introduced 'CONFIG_VENDOR_COREBOOT=y' and was applied before your commit bd328eb to clean up the defconfig. I suspect there was something wrong with 'savedefconfig'?

No, savedefconfig is doing exactly what it should. Before your patch, CONFIG_VENDOR_COREBOOT was the default, explicitly in the Kconfig. Therefore savedefconfig sees it as redundant to specify that in the defconfig as well, so it removed it. When you change that explicit default to something else, it is up to you to change the defconfigs of the old and new default boards.

Quoted 9 lines
Quoted 5 lines

Your other option is to stop defining a default in the Kconfig and instead mark the choice as "optional" (like I did for many other selections like this that had no default explicitly - Kconfig otherwise treats the first entry as default in that case) in which case all defconfigs must have a specified vendor.

OK I've squashed that in and pushed to u-boot-x86/testing. If it looks OK I'll pull it into master.

I just found u-boot.rom created from u-boot-x86/testing branch does not have working network any more on QEMU. It is working on the u-boot-x86/master branch.

=> set serverip 10.10.0.100;set ipaddr 10.10.0.108;set netmask 255.255.255.0;set gatewayip 10.10.0.100
=> run ramboot
*** ERROR: `serverip' not set
*** ERROR: `serverip' not set

Quoted 7 lines
Quoted 4 lines

But 'print' says the env indeed has the 'serverip'. I think there should be something in the upstream that caused this after the u-boot-x86/testing branch rebase. Do you know where might be the problem?

I'll hold off pushing to master for now.

You may need to wait for Joe to submit a patch to fix this networking issue. Please see
http://u-boot.10912.n7.nabble.com/PATCH-imx-missing-CONFIG-NET-after-consolidation-patches-td215762.html
for the discussion.

Regards,
Bin

A message that is not in the archive (sent off-list or not yet received): [email protected]

author
#16 lore raw

PATCH 1/1 x86: Fix regression build issue of coreboot-x86_defconfig

Hi Andy,

Quoted 20 lines · Andy Pont

Hi Bin,

Quoted 10 lines

I just found u-boot.rom created from u-boot-x86/testing branch does not have working network any more on QEMU. It is working on the u-boot-x86/master branch.

=> set serverip 10.10.0.100;set ipaddr 10.10.0.108;set netmask 255.255.255.0;set gatewayip 10.10.0.100
=> run ramboot
*** ERROR: `serverip' not set
*** ERROR: `serverip' not set

Quoted 4 lines

But 'print' says the env indeed has the 'serverip'. I think there should be something in the upstream that caused this after the u-boot-x86/testing branch rebase. Do you know where might be the problem?

This looks very similar to an issue that got patched on the i.MX6 yesterday by Stefano which said:

Quoted 7 lines

commit fd3056337e6fcc140f400e11edd33f6f1cb37de1
"Use env callbacks for net variables" has a side effect on i.MX6 boards because they do not set CONFIG_NET: the ip address results not set, but it is stored in the environment."

Might be something to check...

Thanks for the hint! Indeed there are something wrong in the upstream which broke the networking on several x86 boards.

Regards,
Bin