All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH] cross-canadian/libgcc-common: Fixes for arm multilib
@ 2016-09-28 14:59 Richard Purdie
  2016-09-28 15:28 ` Khem Raj
  0 siblings, 1 reply; 2+ messages in thread
From: Richard Purdie @ 2016-09-28 14:59 UTC (permalink / raw)
  To: openembedded-core

Arm is unusual in that we force it to "linux-gnueabi" and "linux" doesn't
build. This was causing problems for multilib configurations which were assuming
"linux" was the default compiler rather than linux-gnueabi.

This change does two things, ensures symlinks are generated for linux-gnueabi
and also adapts the libgcc code to account for the difference on arm.

It still needs to immediately expand/save TARGET_VENDOR but we defer
deciding what TARGET_OS should be until we know TARGET_ARCH (which the
multilib code may change).

[YOCTO #8642]

Note that sanity tests of a 32 bit arm multilib still break due to issues
with the kernel headers on a mixed bit system. This looks to be a general
headers issue for the platform though and a different type of bug.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

diff --git a/meta/classes/cross-canadian.bbclass b/meta/classes/cross-canadian.bbclass
index 099c0da..21921b3 100644
--- a/meta/classes/cross-canadian.bbclass
+++ b/meta/classes/cross-canadian.bbclass
@@ -80,7 +80,7 @@ python () {
         for extraos in d.getVar("BASECANADIANEXTRAOS", True).split():
             d.appendVar("CANADIANEXTRAOS", " " + extraos + "n32")
     if tarch == "arm" or tarch == "armeb":
-        d.appendVar("CANADIANEXTRAOS", " linux-musleabi linux-uclibceabi")
+        d.appendVar("CANADIANEXTRAOS", " linux-gnueabi linux-musleabi linux-uclibceabi")
         d.setVar("TARGET_OS", "linux-gnueabi")
     else:
         d.setVar("TARGET_OS", "linux")
diff --git a/meta/recipes-devtools/gcc/libgcc-common.inc b/meta/recipes-devtools/gcc/libgcc-common.inc
index 4e4f0fd..1549eb4 100644
--- a/meta/recipes-devtools/gcc/libgcc-common.inc
+++ b/meta/recipes-devtools/gcc/libgcc-common.inc
@@ -123,15 +123,18 @@ fakeroot python do_multilib_install() {
         os.symlink(src, dest)
 }
 
-def get_original_vendoros(d):
-    vendoros = d.expand('${TARGET_VENDOR}-${TARGET_OS}')
+def get_original_os(d):
+    vendoros = d.expand('${TARGET_ARCH}${ORIG_TARGET_VENDOR}-${TARGET_OS}')
     for suffix in [d.getVar('ABIEXTENSION', True), d.getVar('LIBCEXTENSION', True)]:
         if suffix and vendoros.endswith(suffix):
             vendoros = vendoros[:-len(suffix)]
+    # Arm must use linux-gnueabi not linux as only the former is accepted by gcc
+    if vendoros.startswith("arm-") and not vendoros.endswith("-gnueabi"):
+        vendoros = vendoros + "-gnueabi"
     return vendoros
 
-ORIG_TARGET_VENDOROS := "${@get_original_vendoros(d)}"
-BASETARGET_SYS = "${TARGET_ARCH}${ORIG_TARGET_VENDOROS}"
+ORIG_TARGET_VENDOR := "${TARGET_VENDOR}"
+BASETARGET_SYS = "${@get_original_os(d)}"
 
 addtask extra_symlinks after do_multilib_install before do_package do_populate_sysroot
 fakeroot python do_extra_symlinks() {




^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH] cross-canadian/libgcc-common: Fixes for arm multilib
  2016-09-28 14:59 [PATCH] cross-canadian/libgcc-common: Fixes for arm multilib Richard Purdie
@ 2016-09-28 15:28 ` Khem Raj
  0 siblings, 0 replies; 2+ messages in thread
From: Khem Raj @ 2016-09-28 15:28 UTC (permalink / raw)
  To: Richard Purdie; +Cc: Patches and discussions about the oe-core layer

[-- Attachment #1: Type: text/plain, Size: 3279 bytes --]

Seems fine

On Sep 28, 2016 8:00 AM, "Richard Purdie" <
richard.purdie@linuxfoundation.org> wrote:

> Arm is unusual in that we force it to "linux-gnueabi" and "linux" doesn't
> build. This was causing problems for multilib configurations which were
> assuming
> "linux" was the default compiler rather than linux-gnueabi.
>
> This change does two things, ensures symlinks are generated for
> linux-gnueabi
> and also adapts the libgcc code to account for the difference on arm.
>
> It still needs to immediately expand/save TARGET_VENDOR but we defer
> deciding what TARGET_OS should be until we know TARGET_ARCH (which the
> multilib code may change).
>
> [YOCTO #8642]
>
> Note that sanity tests of a 32 bit arm multilib still break due to issues
> with the kernel headers on a mixed bit system. This looks to be a general
> headers issue for the platform though and a different type of bug.
>
> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
>
> diff --git a/meta/classes/cross-canadian.bbclass
> b/meta/classes/cross-canadian.bbclass
> index 099c0da..21921b3 100644
> --- a/meta/classes/cross-canadian.bbclass
> +++ b/meta/classes/cross-canadian.bbclass
> @@ -80,7 +80,7 @@ python () {
>          for extraos in d.getVar("BASECANADIANEXTRAOS", True).split():
>              d.appendVar("CANADIANEXTRAOS", " " + extraos + "n32")
>      if tarch == "arm" or tarch == "armeb":
> -        d.appendVar("CANADIANEXTRAOS", " linux-musleabi
> linux-uclibceabi")
> +        d.appendVar("CANADIANEXTRAOS", " linux-gnueabi linux-musleabi
> linux-uclibceabi")
>          d.setVar("TARGET_OS", "linux-gnueabi")
>      else:
>          d.setVar("TARGET_OS", "linux")
> diff --git a/meta/recipes-devtools/gcc/libgcc-common.inc
> b/meta/recipes-devtools/gcc/libgcc-common.inc
> index 4e4f0fd..1549eb4 100644
> --- a/meta/recipes-devtools/gcc/libgcc-common.inc
> +++ b/meta/recipes-devtools/gcc/libgcc-common.inc
> @@ -123,15 +123,18 @@ fakeroot python do_multilib_install() {
>          os.symlink(src, dest)
>  }
>
> -def get_original_vendoros(d):
> -    vendoros = d.expand('${TARGET_VENDOR}-${TARGET_OS}')
> +def get_original_os(d):
> +    vendoros = d.expand('${TARGET_ARCH}${ORIG_TARGET_VENDOR}-${TARGET_
> OS}')
>      for suffix in [d.getVar('ABIEXTENSION', True),
> d.getVar('LIBCEXTENSION', True)]:
>          if suffix and vendoros.endswith(suffix):
>              vendoros = vendoros[:-len(suffix)]
> +    # Arm must use linux-gnueabi not linux as only the former is accepted
> by gcc
> +    if vendoros.startswith("arm-") and not vendoros.endswith("-gnueabi"):
> +        vendoros = vendoros + "-gnueabi"
>      return vendoros
>
> -ORIG_TARGET_VENDOROS := "${@get_original_vendoros(d)}"
> -BASETARGET_SYS = "${TARGET_ARCH}${ORIG_TARGET_VENDOROS}"
> +ORIG_TARGET_VENDOR := "${TARGET_VENDOR}"
> +BASETARGET_SYS = "${@get_original_os(d)}"
>
>  addtask extra_symlinks after do_multilib_install before do_package
> do_populate_sysroot
>  fakeroot python do_extra_symlinks() {
>
>
> --
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core
>

[-- Attachment #2: Type: text/html, Size: 4432 bytes --]

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2016-09-28 15:28 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-09-28 14:59 [PATCH] cross-canadian/libgcc-common: Fixes for arm multilib Richard Purdie
2016-09-28 15:28 ` Khem Raj

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.