[PATCH 0/2] fix multilib + iso

[PATCH 0/2] fix multilib + iso

[Robert Yang]

The following changes since commit a24c10b7bdab8aa960fdd3a58d2009f24344e579:

  populate_sdk_ext: Use prebuilt uninative tarball (2017-12-11 21:46:33 +0000)

are available in the git repository at:

  git://git.openembedded.org/openembedded-core-contrib rbt/ml_iso
  http://cgit.openembedded.org/openembedded-core-contrib/log/?h=rbt/ml_iso

Robert Yang (2):
  image-live.bbclass: add MLPREFIX to core-image-minimal-initramfs
  image-live.bbclass: remove MLPREFIX from syslinux

 meta/classes/image-live.bbclass                          | 4 ++--
 meta/classes/syslinux.bbclass                            | 2 +-
 meta/recipes-core/images/core-image-minimal-initramfs.bb | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)

-- 
2.7.4

[PATCH 1/2] image-live.bbclass: add MLPREFIX to core-image-minimal-initramfs

[Robert Yang]

The 32bit core-image-minimal-initramfs should be built when build lib32-iso, e.g.:
MACHINE = "qemux86-64"
require conf/multilib.conf
MULTILIBS = "multilib:lib32"
DEFAULTTUNE_virtclass-multilib-lib32 = "x86"
IMAGE_FSTYPES += "iso"

$ bitbake lib32-core-image-minimal
Exception: FileExistsError: [Errno 17] File exists: '/path/to/tmp/sysroots-components/core2-64/qemuwrapper-cross/usr/bin/crossscripts/qemuwrapper' -> '/path/to/tmp/work/qemux86_64-pokymllib32-linux/lib32-core-image-minimal/1.0-r0/lib32-recipe-sysroot/usr/bin/crossscripts/qemuwrapper'

This was because:
lib32-core-image-minimal -> core-image-minimal-initramfs
core-image-minimal-initramfs -> qemuwrapper-cross
lib32-core-image-minimal -> lib32-qemuwrapper-cross

So we got the error, build lib32-core-image-minimal-initramfs can fix the problem.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
---
 meta/classes/image-live.bbclass                          | 2 +-
 meta/recipes-core/images/core-image-minimal-initramfs.bb | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/meta/classes/image-live.bbclass b/meta/classes/image-live.bbclass
index 7a388d5..9f8ff79 100644
--- a/meta/classes/image-live.bbclass
+++ b/meta/classes/image-live.bbclass
@@ -40,7 +40,7 @@ do_bootimg[depends] += "dosfstools-native:do_populate_sysroot \
 
 LABELS_LIVE ?= "boot install"
 ROOT_LIVE ?= "root=/dev/ram0"
-INITRD_IMAGE_LIVE ?= "core-image-minimal-initramfs"
+INITRD_IMAGE_LIVE ?= "${MLPREFIX}core-image-minimal-initramfs"
 INITRD_LIVE ?= "${DEPLOY_DIR_IMAGE}/${INITRD_IMAGE_LIVE}-${MACHINE}.cpio.gz"
 
 LIVE_ROOTFS_TYPE ?= "ext4"
diff --git a/meta/recipes-core/images/core-image-minimal-initramfs.bb b/meta/recipes-core/images/core-image-minimal-initramfs.bb
index 5794a25..c446e87 100644
--- a/meta/recipes-core/images/core-image-minimal-initramfs.bb
+++ b/meta/recipes-core/images/core-image-minimal-initramfs.bb
@@ -8,7 +8,7 @@ PACKAGE_INSTALL = "initramfs-live-boot initramfs-live-install initramfs-live-ins
 # Do not pollute the initrd image with rootfs features
 IMAGE_FEATURES = ""
 
-export IMAGE_BASENAME = "core-image-minimal-initramfs"
+export IMAGE_BASENAME = "${MLPREFIX}core-image-minimal-initramfs"
 IMAGE_LINGUAS = ""
 
 LICENSE = "MIT"
-- 
2.7.4

[PATCH 2/2] image-live.bbclass: remove MLPREFIX from syslinux

[Robert Yang]

Fixed:
MACHINE = "qemux86-64"
require conf/multilib.conf
MULTILIBS = "multilib:lib32"
DEFAULTTUNE_virtclass-multilib-lib32 = "x86"
IMAGE_FSTYPES += "iso"

$ bitbake lib32-core-image-minimal
ERROR: lib32-core-image-minimal-1.0-r0 do_bootimg: The file /usr/include/printf.h is installed by both glibc and lib32-glibc, aborting

This was because:
lib32-syslinux -> lib32-glibc
virtual/kernel -> glibc

We can build 64bit syslinux (only build, not install) to fix the problem, the
do_bootimg only needs several data files of syslinux such as vesamenu.c32,
these files are not arch related.

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
---
 meta/classes/image-live.bbclass | 2 +-
 meta/classes/syslinux.bbclass   | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/meta/classes/image-live.bbclass b/meta/classes/image-live.bbclass
index 9f8ff79..8872195 100644
--- a/meta/classes/image-live.bbclass
+++ b/meta/classes/image-live.bbclass
@@ -31,7 +31,7 @@ do_bootimg[depends] += "dosfstools-native:do_populate_sysroot \
                         mtools-native:do_populate_sysroot \
                         cdrtools-native:do_populate_sysroot \
                         virtual/kernel:do_deploy \
-                        ${MLPREFIX}syslinux:do_populate_sysroot \
+                        syslinux:do_populate_sysroot \
                         syslinux-native:do_populate_sysroot \
                         ${@oe.utils.ifelse(d.getVar('COMPRESSISO', False),'zisofs-tools-native:do_populate_sysroot','')} \
                         ${PN}:do_image_${@d.getVar('LIVE_ROOTFS_TYPE').replace('-', '_')} \
diff --git a/meta/classes/syslinux.bbclass b/meta/classes/syslinux.bbclass
index d6f8824..98939ba 100644
--- a/meta/classes/syslinux.bbclass
+++ b/meta/classes/syslinux.bbclass
@@ -17,7 +17,7 @@
 # ${SYSLINUX_SERIAL_TTY} - Set alternate console=tty... kernel boot argument
 # ${SYSLINUX_KERNEL_ARGS} - Add additional kernel arguments
 
-do_bootimg[depends] += "${MLPREFIX}syslinux:do_populate_sysroot \
+do_bootimg[depends] += "syslinux:do_populate_sysroot \
                         syslinux-native:do_populate_sysroot"
 
 ISOLINUXDIR ?= "/isolinux"
-- 
2.7.4

Re: [PATCH 2/2] image-live.bbclass: remove MLPREFIX from syslinux

[Richard Purdie]

On Wed, 2017-12-13 at 10:45 +0800, Robert Yang wrote:
> Fixed:
> MACHINE = "qemux86-64"
> require conf/multilib.conf
> MULTILIBS = "multilib:lib32"
> DEFAULTTUNE_virtclass-multilib-lib32 = "x86"
> IMAGE_FSTYPES += "iso"
> 
> $ bitbake lib32-core-image-minimal
> ERROR: lib32-core-image-minimal-1.0-r0 do_bootimg: The file
> /usr/include/printf.h is installed by both glibc and lib32-glibc,
> aborting
> 
> This was because:
> lib32-syslinux -> lib32-glibc
> virtual/kernel -> glibc
> 
> We can build 64bit syslinux (only build, not install) to fix the
> problem, the
> do_bootimg only needs several data files of syslinux such as
> vesamenu.c32,
> these files are not arch related.

Hi Robert,

I've been thinking about this one and I'm not 100% convinced this is
the right thing to do.

When we build "lib32-core-image-minimal", one of the things we want to
avoid is building two different toolchains, there should only be one
needed for this image.

If there is a dependency on "syslinux", that will need the non-multilib 
toolchain. I suspect that is why libX-syslinux was used as a
dependency.

Also, there are some things which never make sense as a multilib, the
kernel is one example and I'm starting to wonder if syslinux would be
another. In the kernel (and kernel module) case we'd provide all the
libX variants from the same recipe, we may want to do that for
syslinux.

It may be we can't avoid the multiple compiler issue and the current
codebase may not do so, I think currently we can avoid multiple glibc
though.

Regardless, I do think this needs a little more thought, we also need
better multiple test cases as we're not catching issues like this. I
think this needs to be revisited along with your outstanding multilib
patch series which I haven't found the time to review yet (sorry).

Cheers,

Richard

Re: [PATCH 2/2] image-live.bbclass: remove MLPREFIX from syslinux

[Robert Yang]

On 01/03/2018 08:43 PM, Richard Purdie wrote:
> On Wed, 2017-12-13 at 10:45 +0800, Robert Yang wrote:
>> Fixed:
>> MACHINE = "qemux86-64"
>> require conf/multilib.conf
>> MULTILIBS = "multilib:lib32"
>> DEFAULTTUNE_virtclass-multilib-lib32 = "x86"
>> IMAGE_FSTYPES += "iso"
>>
>> $ bitbake lib32-core-image-minimal
>> ERROR: lib32-core-image-minimal-1.0-r0 do_bootimg: The file
>> /usr/include/printf.h is installed by both glibc and lib32-glibc,
>> aborting
>>
>> This was because:
>> lib32-syslinux -> lib32-glibc
>> virtual/kernel -> glibc
>>
>> We can build 64bit syslinux (only build, not install) to fix the
>> problem, the
>> do_bootimg only needs several data files of syslinux such as
>> vesamenu.c32,
>> these files are not arch related.
> 
> Hi Robert,

Hi RP,

Thanks for the reply.

> 
> I've been thinking about this one and I'm not 100% convinced this is
> the right thing to do.
> 
> When we build "lib32-core-image-minimal", one of the things we want to
> avoid is building two different toolchains, there should only be one
> needed for this image.
> 
> If there is a dependency on "syslinux", that will need the non-multilib
> toolchain. I suspect that is why libX-syslinux was used as a
> dependency.
> 
> Also, there are some things which never make sense as a multilib, the
> kernel is one example and I'm starting to wonder if syslinux would be
> another. In the kernel (and kernel module) case we'd provide all the
> libX variants from the same recipe, we may want to do that for
> syslinux.

I think that syslinux is different from kernel, 64bit kernel can provide
32bit kernel via kernel config, but syslinux can't (except these non
arch files).

> 
> It may be we can't avoid the multiple compiler issue and the current
> codebase may not do so, I think currently we can avoid multiple glibc
> though.

I'm not sure about a problem on this, should lib32-image can run 64bit
programs or not ? If yes, then kernel should be 64bit, and we can't avoid
building 64bit compilers. And I'm leaning to yes since we call it multilib,
the pure 32bit image which can't run 64bit programs can't be called multilib.

I did a rough search on why glibc is built, when bitbake lib32-image, it
is because:

64 bit kernel -> gcc-cross-x86_64 -> virtual/${TARGET_PREFIX}libc-for-gcc,
and the TARGET_PREFIX is x86_64 when building gcc-cross-x86_64,
While the virtual/x86_64-poky-linux-libc-for-gcc is provided by glibc.

I tried to remove the depends, both "bitbake gcc-cross-x86_64" and
"bitbake linux-yocto" can be built, but other recipes such as quilt would
be failed:

| 
/workspace1/lyang1/test_1/tmp/work/core2-64-poky-linux/libgcc/7.2.0-r0/recipe-sysroot-native/usr/bin/x86_64-poky-linux/../../libexec/x86_64-poky-linux/gcc/x86_64-poky-linux/7.2.0/ld: 
cannot find crti.o: No such file or directory
| 
/workspace1/lyang1/test_1/tmp/work/core2-64-poky-linux/libgcc/7.2.0-r0/recipe-sysroot-native/usr/bin/x86_64-poky-linux/../../libexec/x86_64-poky-linux/gcc/x86_64-poky-linux/7.2.0/ld: 
cannot find -lc
| 
/workspace1/lyang1/test_1/tmp/work/core2-64-poky-linux/libgcc/7.2.0-r0/recipe-sysroot-native/usr/bin/x86_64-poky-linux/../../libexec/x86_64-poky-linux/gcc/x86_64-poky-linux/7.2.0/ld: 
cannot find crtn.o: No such file or directory
| collect2: error: ld returned 1 exit status
| Makefile:978: recipe for target 'libgcc_s.so' failed

And move virtual/${TARGET_PREFIX}libc-for-gcc to BASE_DEFAULT_DEPS can't make
it work either, so it seems that we can't avoid building glibc.

> 
> Regardless, I do think this needs a little more thought, we also need
> better multiple test cases as we're not catching issues like this. 
> think this needs to be revisited along with your outstanding multilib
> patch series which I haven't found the time to review yet (sorry).

That's all right, I'm very glad to make mutilib work well, including
adding test cases for them. It is nearly broken after changed from
smart + rpm5 to dnf + rpm4, those patches fixed the problem.

// Robert

> 
> Cheers,
> 
> Richard
> 

Re: [PATCH 2/2] image-live.bbclass: remove MLPREFIX from syslinux

[Richard Purdie]

On Wed, 2018-01-03 at 21:56 +0800, Robert Yang wrote:
> > Also, there are some things which never make sense as a multilib,
> > the
> > kernel is one example and I'm starting to wonder if syslinux would
> > be
> > another. In the kernel (and kernel module) case we'd provide all
> > the
> > libX variants from the same recipe, we may want to do that for
> > syslinux.
> I think that syslinux is different from kernel, 64bit kernel can
> provide 32bit kernel via kernel config, but syslinux can't (except
> these non arch files).

Think about this differently. The system can ever only boot one single
kernel. The image can have several different multilibs running at once
in different userspace processes but there can only ever be one running
kernel.

How that kernel is configured is obviously important but the key thing
is there can only be one running.

The bootloader is similar in that you likely only ever want one and I
suspect syslinux is similar.

Also, don't confuse this with multi-boot or multi partition systems
where there could be a "main" and a "backup" kernel. In those cases
there would only ever be one running at once.

> > It may be we can't avoid the multiple compiler issue and the
> > current
> > codebase may not do so, I think currently we can avoid multiple
> > glibc
> > though.
> I'm not sure about a problem on this, should lib32-image can run
> 64bit programs or not ? If yes, then kernel should be 64bit, and we
> can't avoid building 64bit compilers. And I'm leaning to yes since we
> call it multilib, the pure 32bit image which can't run 64bit programs
> can't be called multilib.

That is a configuration issue for the multilib you select. In general,
yes you'd want a 64 bit kernel.

> I did a rough search on why glibc is built, when bitbake lib32-image, 
> it is because:
> 
> 64 bit kernel -> gcc-cross-x86_64 -> virtual/${TARGET_PREFIX}libc-
> for-gcc,
> and the TARGET_PREFIX is x86_64 when building gcc-cross-x86_64,
> While the virtual/x86_64-poky-linux-libc-for-gcc is provided by
> glibc.
> 
> I tried to remove the depends, both "bitbake gcc-cross-x86_64" and
> "bitbake linux-yocto" can be built, but other recipes such as quilt
> would be failed:

I find it interesting that gcc-cross-x86_64 would build without glibc.
If that really is the case we may be able to speed up our compiler
bootstrap so that could be worth investigating further.

Its not surprising that libgcc won't build without glibc though.
Perhaps gcc-cross works since we split out the build of libgcc?

> > collect2: error: ld returned 1 exit status
> > Makefile:978: recipe for target 'libgcc_s.so' failed
> And move virtual/${TARGET_PREFIX}libc-for-gcc to BASE_DEFAULT_DEPS
> can't make it work either, so it seems that we can't avoid building
> glibc.

No, I'd forgotten that gcc-cross depends on glibc. Based on the above
there may still be some optimisation we can make here.

In the back of my mind, I'm concious that one x86 cross compiler should
be able to work for 32 and 64 bit too, its just the compiler options
and libgcc etc. would need to be generated for both...

> > Regardless, I do think this needs a little more thought, we also
> > need
> > better multiple test cases as we're not catching issues like this. 
> > think this needs to be revisited along with your outstanding
> > multilib
> > patch series which I haven't found the time to review yet (sorry).
> That's all right, I'm very glad to make mutilib work well, including
> adding test cases for them. It is nearly broken after changed from
> smart + rpm5 to dnf + rpm4, those patches fixed the problem.

Agreed, its important and on my list of things to review, I've just had
to focus on getting build testing working properly and now I can try
and clear some of the patch backlog.

Cheers,

Richard

Re: [PATCH 2/2] image-live.bbclass: remove MLPREFIX from syslinux

[Robert Yang]

Hi RP,

Thanks for the reply.

On 01/03/2018 10:19 PM, Richard Purdie wrote:
> On Wed, 2018-01-03 at 21:56 +0800, Robert Yang wrote:
>>> Also, there are some things which never make sense as a multilib,
>>> the
>>> kernel is one example and I'm starting to wonder if syslinux would
>>> be
>>> another. In the kernel (and kernel module) case we'd provide all
>>> the
>>> libX variants from the same recipe, we may want to do that for
>>> syslinux.
>> I think that syslinux is different from kernel, 64bit kernel can
>> provide 32bit kernel via kernel config, but syslinux can't (except
>> these non arch files).
> 
> Think about this differently. The system can ever only boot one single
> kernel. The image can have several different multilibs running at once
> in different userspace processes but there can only ever be one running
> kernel.
> 
> How that kernel is configured is obviously important but the key thing
> is there can only be one running.
> 
> The bootloader is similar in that you likely only ever want one and I
> suspect syslinux is similar.

Makes sense, I will work on it.

> 
> Also, don't confuse this with multi-boot or multi partition systems
> where there could be a "main" and a "backup" kernel. In those cases
> there would only ever be one running at once.
> 
>>> It may be we can't avoid the multiple compiler issue and the
>>> current
>>> codebase may not do so, I think currently we can avoid multiple
>>> glibc
>>> though.
>> I'm not sure about a problem on this, should lib32-image can run
>> 64bit programs or not ? If yes, then kernel should be 64bit, and we
>> can't avoid building 64bit compilers. And I'm leaning to yes since we
>> call it multilib, the pure 32bit image which can't run 64bit programs
>> can't be called multilib.
> 
> That is a configuration issue for the multilib you select. In general,
> yes you'd want a 64 bit kernel.
> 
>> I did a rough search on why glibc is built, when bitbake lib32-image,
>> it is because:
>>
>> 64 bit kernel -> gcc-cross-x86_64 -> virtual/${TARGET_PREFIX}libc-
>> for-gcc,
>> and the TARGET_PREFIX is x86_64 when building gcc-cross-x86_64,
>> While the virtual/x86_64-poky-linux-libc-for-gcc is provided by
>> glibc.
>>
>> I tried to remove the depends, both "bitbake gcc-cross-x86_64" and
>> "bitbake linux-yocto" can be built, but other recipes such as quilt
>> would be failed:
> 
> I find it interesting that gcc-cross-x86_64 would build without glibc.
> If that really is the case we may be able to speed up our compiler
> bootstrap so that could be worth investigating further.
> 
> Its not surprising that libgcc won't build without glibc though.
> Perhaps gcc-cross works since we split out the build of libgcc?
> 
>>> collect2: error: ld returned 1 exit status
>>> Makefile:978: recipe for target 'libgcc_s.so' failed
>> And move virtual/${TARGET_PREFIX}libc-for-gcc to BASE_DEFAULT_DEPS
>> can't make it work either, so it seems that we can't avoid building
>> glibc.
> 
> No, I'd forgotten that gcc-cross depends on glibc. Based on the above
> there may still be some optimisation we can make here.
> 
> In the back of my mind, I'm concious that one x86 cross compiler should
> be able to work for 32 and 64 bit too, its just the compiler options
> and libgcc etc. would need to be generated for both...

Build one toolchain for both x86 32 and 64 is possible, I'd like to work
on it if it is worth.

// Robert

> 
>>> Regardless, I do think this needs a little more thought, we also
>>> need
>>> better multiple test cases as we're not catching issues like this.
>>> think this needs to be revisited along with your outstanding
>>> multilib
>>> patch series which I haven't found the time to review yet (sorry).
>> That's all right, I'm very glad to make mutilib work well, including
>> adding test cases for them. It is nearly broken after changed from
>> smart + rpm5 to dnf + rpm4, those patches fixed the problem.
> 
> Agreed, its important and on my list of things to review, I've just had
> to focus on getting build testing working properly and now I can try
> and clear some of the patch backlog.
> 
> Cheers,
> 
> Richard
> 

Re: [PATCH 2/2] image-live.bbclass: remove MLPREFIX from syslinux

[Robert Yang]

Hi RP,

On 01/04/2018 07:13 PM, Robert Yang wrote:
> Hi RP,
> 
> Thanks for the reply.
> 
> On 01/03/2018 10:19 PM, Richard Purdie wrote:
>> On Wed, 2018-01-03 at 21:56 +0800, Robert Yang wrote:
>>>> Also, there are some things which never make sense as a multilib,
>>>> the
>>>> kernel is one example and I'm starting to wonder if syslinux would
>>>> be
>>>> another. In the kernel (and kernel module) case we'd provide all
>>>> the
>>>> libX variants from the same recipe, we may want to do that for
>>>> syslinux.
>>> I think that syslinux is different from kernel, 64bit kernel can
>>> provide 32bit kernel via kernel config, but syslinux can't (except
>>> these non arch files).
>>
>> Think about this differently. The system can ever only boot one single
>> kernel. The image can have several different multilibs running at once
>> in different userspace processes but there can only ever be one running
>> kernel.
>>
>> How that kernel is configured is obviously important but the key thing
>> is there can only be one running.
>>
>> The bootloader is similar in that you likely only ever want one and I
>> suspect syslinux is similar.
> 
> Makes sense, I will work on it.

I tried syslinux, we can make syslinux provide lib32-syslinux, if we treat
it as a bootloader, it is OK since we use syslinux-native to install the
bootloader, but if we treat it as an application, e.g.:

IMAGE_INSTALL += "syslinux"
And then:
bitbake lib32-core-image-minimal

I think that we need a 32bit syslinux to be installed rather than 64bit,
so I'm afraid that we can't let syslinux provide lib32-syslinux.

// Robert

> 
>>
>> Also, don't confuse this with multi-boot or multi partition systems
>> where there could be a "main" and a "backup" kernel. In those cases
>> there would only ever be one running at once.
>>
>>>> It may be we can't avoid the multiple compiler issue and the
>>>> current
>>>> codebase may not do so, I think currently we can avoid multiple
>>>> glibc
>>>> though.
>>> I'm not sure about a problem on this, should lib32-image can run
>>> 64bit programs or not ? If yes, then kernel should be 64bit, and we
>>> can't avoid building 64bit compilers. And I'm leaning to yes since we
>>> call it multilib, the pure 32bit image which can't run 64bit programs
>>> can't be called multilib.
>>
>> That is a configuration issue for the multilib you select. In general,
>> yes you'd want a 64 bit kernel.
>>
>>> I did a rough search on why glibc is built, when bitbake lib32-image,
>>> it is because:
>>>
>>> 64 bit kernel -> gcc-cross-x86_64 -> virtual/${TARGET_PREFIX}libc-
>>> for-gcc,
>>> and the TARGET_PREFIX is x86_64 when building gcc-cross-x86_64,
>>> While the virtual/x86_64-poky-linux-libc-for-gcc is provided by
>>> glibc.
>>>
>>> I tried to remove the depends, both "bitbake gcc-cross-x86_64" and
>>> "bitbake linux-yocto" can be built, but other recipes such as quilt
>>> would be failed:
>>
>> I find it interesting that gcc-cross-x86_64 would build without glibc.
>> If that really is the case we may be able to speed up our compiler
>> bootstrap so that could be worth investigating further.
>>
>> Its not surprising that libgcc won't build without glibc though.
>> Perhaps gcc-cross works since we split out the build of libgcc?
>>
>>>> collect2: error: ld returned 1 exit status
>>>> Makefile:978: recipe for target 'libgcc_s.so' failed
>>> And move virtual/${TARGET_PREFIX}libc-for-gcc to BASE_DEFAULT_DEPS
>>> can't make it work either, so it seems that we can't avoid building
>>> glibc.
>>
>> No, I'd forgotten that gcc-cross depends on glibc. Based on the above
>> there may still be some optimisation we can make here.
>>
>> In the back of my mind, I'm concious that one x86 cross compiler should
>> be able to work for 32 and 64 bit too, its just the compiler options
>> and libgcc etc. would need to be generated for both...
> 
> Build one toolchain for both x86 32 and 64 is possible, I'd like to work
> on it if it is worth.
> 
> // Robert
> 
>>
>>>> Regardless, I do think this needs a little more thought, we also
>>>> need
>>>> better multiple test cases as we're not catching issues like this.
>>>> think this needs to be revisited along with your outstanding
>>>> multilib
>>>> patch series which I haven't found the time to review yet (sorry).
>>> That's all right, I'm very glad to make mutilib work well, including
>>> adding test cases for them. It is nearly broken after changed from
>>> smart + rpm5 to dnf + rpm4, those patches fixed the problem.
>>
>> Agreed, its important and on my list of things to review, I've just had
>> to focus on getting build testing working properly and now I can try
>> and clear some of the patch backlog.
>>
>> Cheers,
>>
>> Richard
>>

Re: [PATCH 2/2] image-live.bbclass: remove MLPREFIX from syslinux

[Robert Yang]

On 01/04/2018 07:13 PM, Robert Yang wrote:
> Hi RP,
> 
> Thanks for the reply.
> 
> On 01/03/2018 10:19 PM, Richard Purdie wrote:
>> On Wed, 2018-01-03 at 21:56 +0800, Robert Yang wrote:
>>>> Also, there are some things which never make sense as a multilib,
>>>> the
>>>> kernel is one example and I'm starting to wonder if syslinux would
>>>> be
>>>> another. In the kernel (and kernel module) case we'd provide all
>>>> the
>>>> libX variants from the same recipe, we may want to do that for
>>>> syslinux.
>>> I think that syslinux is different from kernel, 64bit kernel can
>>> provide 32bit kernel via kernel config, but syslinux can't (except
>>> these non arch files).
>>
>> Think about this differently. The system can ever only boot one single
>> kernel. The image can have several different multilibs running at once
>> in different userspace processes but there can only ever be one running
>> kernel.
>>
>> How that kernel is configured is obviously important but the key thing
>> is there can only be one running.
>>
>> The bootloader is similar in that you likely only ever want one and I
>> suspect syslinux is similar.
> 
> Makes sense, I will work on it.
> 
>>
>> Also, don't confuse this with multi-boot or multi partition systems
>> where there could be a "main" and a "backup" kernel. In those cases
>> there would only ever be one running at once.
>>
>>>> It may be we can't avoid the multiple compiler issue and the
>>>> current
>>>> codebase may not do so, I think currently we can avoid multiple
>>>> glibc
>>>> though.
>>> I'm not sure about a problem on this, should lib32-image can run
>>> 64bit programs or not ? If yes, then kernel should be 64bit, and we
>>> can't avoid building 64bit compilers. And I'm leaning to yes since we
>>> call it multilib, the pure 32bit image which can't run 64bit programs
>>> can't be called multilib.
>>
>> That is a configuration issue for the multilib you select. In general,
>> yes you'd want a 64 bit kernel.
>>
>>> I did a rough search on why glibc is built, when bitbake lib32-image,
>>> it is because:
>>>
>>> 64 bit kernel -> gcc-cross-x86_64 -> virtual/${TARGET_PREFIX}libc-
>>> for-gcc,
>>> and the TARGET_PREFIX is x86_64 when building gcc-cross-x86_64,
>>> While the virtual/x86_64-poky-linux-libc-for-gcc is provided by
>>> glibc.
>>>
>>> I tried to remove the depends, both "bitbake gcc-cross-x86_64" and
>>> "bitbake linux-yocto" can be built, but other recipes such as quilt
>>> would be failed:
>>
>> I find it interesting that gcc-cross-x86_64 would build without glibc.
>> If that really is the case we may be able to speed up our compiler
>> bootstrap so that could be worth investigating further.
>>
>> Its not surprising that libgcc won't build without glibc though.
>> Perhaps gcc-cross works since we split out the build of libgcc?
>>
>>>> collect2: error: ld returned 1 exit status
>>>> Makefile:978: recipe for target 'libgcc_s.so' failed
>>> And move virtual/${TARGET_PREFIX}libc-for-gcc to BASE_DEFAULT_DEPS
>>> can't make it work either, so it seems that we can't avoid building
>>> glibc.
>>
>> No, I'd forgotten that gcc-cross depends on glibc. Based on the above
>> there may still be some optimisation we can make here.
>>
>> In the back of my mind, I'm concious that one x86 cross compiler should
>> be able to work for 32 and 64 bit too, its just the compiler options
>> and libgcc etc. would need to be generated for both...
> 
> Build one toolchain for both x86 32 and 64 is possible, I'd like to work
> on it if it is worth.

Add Mark in the loop, here is his comments:
"""
We had a single toolchain built outside of the Yocto Project environment with a
custom spec file.  The only way to do this inside of the Yocto Project would be
to generate a custom spec file for the cross-toolchain based on a specific
configuration.

In the past when this was explored, it was deemed to not be a 'generic'
solution.  So using the stock GCC spec file(s) and/or multiple toolchains was
the better generic approach.
"""

So I'm leaning to keep the current status, multiple toolchains for multilib.

// Robert

> 
> // Robert
> 
>>
>>>> Regardless, I do think this needs a little more thought, we also
>>>> need
>>>> better multiple test cases as we're not catching issues like this.
>>>> think this needs to be revisited along with your outstanding
>>>> multilib
>>>> patch series which I haven't found the time to review yet (sorry).
>>> That's all right, I'm very glad to make mutilib work well, including
>>> adding test cases for them. It is nearly broken after changed from
>>> smart + rpm5 to dnf + rpm4, those patches fixed the problem.
>>
>> Agreed, its important and on my list of things to review, I've just had
>> to focus on getting build testing working properly and now I can try
>> and clear some of the patch backlog.
>>
>> Cheers,
>>
>> Richard
>>