Skip to contentSkip to conversation

qemux86: use a Core 2 Duo CPU instead of the original circa-1993 Pentium

PATCH In review Alexander Kanavin 5 participants 12 messages May 13, 2019 – May 14, 2019

lore
Share & tools
lore
#0 lore raw

PATCH 1/2 qemux86: use a Core 2 Duo CPU instead of the original circa-1993 Pentium

This matches what the qemux86_64 is currently using, and will allow testing the instructions added in the meantime; particularly various SSE extensions are now enabled.

---
 meta/conf/machine/include/qemuboot-x86.inc | 4 ++--
 meta/conf/machine/qemux86.conf             | 3 +--
 2 files changed, 3 insertions(+), 4 deletions(-)
meta/conf/machine/include/qemuboot-x86.inc +2 -2
index 5fdbe4df50e..82ce46d8604 100644
@@ -1,8 +1,8 @@
 # For runqemu
 IMAGE_CLASSES += "qemuboot"
 QB_SYSTEM_NAME_x86 = "qemu-system-i386"
-QB_CPU_x86 = "-cpu pentium2"
-QB_CPU_KVM_x86 = "-cpu pentium2"
+QB_CPU_x86 = "-cpu core2duo"
+QB_CPU_KVM_x86 = "-cpu core2duo"
 
 QB_SYSTEM_NAME_x86-64 = "qemu-system-x86_64"
 QB_CPU_x86-64 = "-cpu core2duo"
meta/conf/machine/qemux86.conf +1 -2
index f2434a4b5f5..272ad1e1870 100644
@@ -8,8 +8,7 @@ PREFERRED_PROVIDER_virtual/libgles1 ?= "mesa"
 PREFERRED_PROVIDER_virtual/libgles2 ?= "mesa"
 
 require conf/machine/include/qemu.inc
-DEFAULTTUNE ?= "i586"
-X86ARCH32 ?= "i586"
+DEFAULTTUNE ?= "core2-32"
 require conf/machine/include/tune-corei7.inc
 require conf/machine/include/qemuboot-x86.inc
 
Signature · 1 line
2.17.1
#1 lore raw

PATCH 2/2 packagegroup-self-hosted: drop epiphany

Originally a web browser was added to build-appliance-image to enable Hob's use cases that involve opening a URI; with Hob long gone, a web browser is no longer necessary to have in build-appliance-image.

This will also address the out-of-resources problem when build-appliance-image builds itself in a VM, as a test case.

---
 meta/recipes-core/packagegroups/packagegroup-self-hosted.bb | 1 -
 1 file changed, 1 deletion(-)
meta/recipes-core/packagegroups/packagegroup-self-hosted.bb +0 -1
index b296db5457e..f206cee3731 100644
@@ -189,7 +189,6 @@ RDEPENDS_packagegroup-self-hosted-graphics = "\
     libx11-dev \
     adwaita-icon-theme \
     xdg-utils \
-    epiphany \
     l3afpad \
     pcmanfm \
     vte \
Signature · 1 line
2.17.1
#2 lore raw
In reply to Alexander Kanavin
Quoted 4 lines · Alexander Kanavin

This matches what the qemux86_64 is currently using, and will allow testing the instructions added in the meantime; particularly various SSE extensions are now enabled.

While I like this change I also worry if we are leaving out any significant x86 implementation out there in embedded space x86-64 is relatively newer compared to 32 bit which was long stable

Quoted 47 lines
---
 meta/conf/machine/include/qemuboot-x86.inc | 4 ++--
 meta/conf/machine/qemux86.conf             | 3 +--
 2 files changed, 3 insertions(+), 4 deletions(-)
meta/conf/machine/include/qemuboot-x86.inc +2 -2
index 5fdbe4df50e..82ce46d8604 100644
@@ -1,8 +1,8 @@
 # For runqemu
 IMAGE_CLASSES += "qemuboot"
 QB_SYSTEM_NAME_x86 = "qemu-system-i386"
-QB_CPU_x86 = "-cpu pentium2"
-QB_CPU_KVM_x86 = "-cpu pentium2"
+QB_CPU_x86 = "-cpu core2duo"
+QB_CPU_KVM_x86 = "-cpu core2duo"
 
 QB_SYSTEM_NAME_x86-64 = "qemu-system-x86_64"
 QB_CPU_x86-64 = "-cpu core2duo"
meta/conf/machine/qemux86.conf +1 -3
index f2434a4b5f5..272ad1e1870 100644
@@ -8,8 +8,7 @@ PREFERRED_PROVIDER_virtual/libgles1 ?= "mesa"
 PREFERRED_PROVIDER_virtual/libgles2 ?= "mesa"
 
 require conf/machine/include/qemu.inc
-DEFAULTTUNE ?= "i586"
-X86ARCH32 ?= "i586"
+DEFAULTTUNE ?= "core2-32"
 require conf/machine/include/tune-corei7.inc
 require conf/machine/include/qemuboot-x86.inc
 
--

2.17.1

--
_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.openembedded.org/mailman/listinfo/openembedded-core

#3 lore raw
Quoted 1 line · Khem Raj

While I like this change I also worry if we are leaving out any significant x86 implementation out there in embedded space x86-64 is relatively newer compared to 32 bit which was long stable

core2duo isn't exactly a modern chipset though, first released in 2006. Does anyone *know* of any significant 32-bit x86 deployments which are using significantly older hardware?

Ross

#4 lore raw
Quoted 9 lines · Ross
Quoted 1 line · Khem Raj

While I like this change I also worry if we are leaving out any

significant x86 implementation out there in embedded space x86-64 is relatively newer compared to 32 bit which was long stable

core2duo isn't exactly a modern chipset though, first released in 2006. Does anyone *know* of any significant 32-bit x86 deployments which are using significantly older hardware?

If intel bsp folks are okay with this then let’s go for it

Quoted 3 lines

Ross

#5 lore raw
Quoted 1 line · Khem Raj

If intel bsp folks are okay with this then let’s go for it

Putting my Intel hat on, we don't use these BSPs. meta-intel ships its own 32-bit and 64-bit machines.

Ross

#6 lore raw
In reply to Khem Raj

Note that the 32 bit machine from meta-intel has actually been using core2-32 tune for a while, so we are simply catching up here.

Alex

Quoted 17 lines · Khem Raj
Quoted 7 lines · Ross
Quoted 1 line · Khem Raj

While I like this change I also worry if we are leaving out any significant x86 implementation out there in embedded space x86-64 is relatively newer compared to 32 bit which was long stable

core2duo isn't exactly a modern chipset though, first released in 2006. Does anyone *know* of any significant 32-bit x86 deployments which are using significantly older hardware?

If intel bsp folks are okay with this then let’s go for it

Quoted 3 lines

Ross

#7 lore raw

Sounds good then unless someone else has more
Info

Quoted 27 lines · Alexander Kanavin

Note that the 32 bit machine from meta-intel has actually been using core2-32 tune for a while, so we are simply catching up here.

Alex

Quoted 4 lines · Khem Raj

On Mon, May 13, 2019 at 9:00 AM Burton, Ross <[email protected]>

Quoted 3 lines · wrote:
Quoted 3 lines · Khem Raj
Quoted 1 line

While I like this change I also worry if we are leaving out any

significant x86 implementation out there in embedded space x86-64 is relatively newer compared to 32 bit which was long stable

Quoted 10 lines
Quoted 4 lines

core2duo isn't exactly a modern chipset though, first released in 2006. Does anyone *know* of any significant 32-bit x86 deployments which are using significantly older hardware?

If intel bsp folks are okay with this then let’s go for it

Quoted 3 lines

Ross

#8 lore raw
In reply to Khem Raj
Quoted 10 lines · Khem Raj
Quoted 3 lines · Alexander Kanavin

This matches what the qemux86_64 is currently using, and will allow testing the instructions added in the meantime; particularly various SSE extensions are now enabled.

While I like this change I also worry if we are leaving out any significant x86 implementation out there in embedded space x86-64 is relatively newer compared to 32 bit which was long stable

Intel Quark and AMD Geode LX are still available to buy today (Geode for at least 2 more years).

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#9 lore raw
Quoted 9 lines · Adrian Bunk
Quoted 7 lines · Khem Raj
Quoted 3 lines · Alexander Kanavin

This matches what the qemux86_64 is currently using, and will allow testing the instructions added in the meantime; particularly various SSE extensions are now enabled.

Quoted 6 lines
Quoted 3 lines

While I like this change I also worry if we are leaving out any significant x86 implementation out there in embedded space x86-64 is relatively newer compared to 32 bit which was long stable

Intel Quark and AMD Geode LX are still available to buy today (Geode for at least 2 more years).

I think they have their own BSP layers, so they lose out on least common denominator here. I don't have strong objection to that, given that using core2 is going to enhance our testing coverage quite a bit.

Quoted 11 lines

cu
Adrian

--

"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of rain for many days.
"Only a promise," Lao Er said.
Pearl S. Buck - Dragon Seed

#10 lore raw
In reply to Alexander Kanavin
Quoted 9 lines · Alexander Kanavin

This matches what the qemux86_64 is currently using, and will allow testing the instructions added in the meantime; particularly various SSE extensions are now enabled.

---
 meta/conf/machine/include/qemuboot-x86.inc | 4 ++--
 meta/conf/machine/qemux86.conf             | 3 +--
 2 files changed, 3 insertions(+), 4 deletions(-)

There is some kind of bad assumption somewhere in the sdk/mingw code:

https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/152

and another glitch somewhere in the SDK code:

https://autobuilder.yoctoproject.org/typhoon/#/builders/59/builds/588

I was also worried that:

http://git.yoctoproject.org/cgit.cgi/yocto-autobuilder-helper/tree/scripts/publish-artefacts
may need fixes but it looks like it might be ok.

Cheers,

Richard

#11 lore raw
Quoted 12 lines · Richard Purdie
Quoted 9 lines · Alexander Kanavin

This matches what the qemux86_64 is currently using, and will allow testing the instructions added in the meantime; particularly various SSE extensions are now enabled.

---
 meta/conf/machine/include/qemuboot-x86.inc | 4 ++--
 meta/conf/machine/qemux86.conf             | 3 +--
 2 files changed, 3 insertions(+), 4 deletions(-)
Quoted 7 lines

There is some kind of bad assumption somewhere in the sdk/mingw code:

https://autobuilder.yoctoproject.org/typhoon/#/builders/83/builds/152

and another glitch somewhere in the SDK code:

https://autobuilder.yoctoproject.org/typhoon/#/builders/59/builds/588

Seems like packagegroup-cross-canadian is erroneously taken from sstate, instead of being repackaged when TUNE_ARCH changes but MACHINE does not. I will send a fix in a moment, even though I am not totally sure if it's the correct one.

Alex