populate_sdk_base.bbclass: fixes for multilib sdk

For a clear look when input.

[YOCTO #8089]

Signed-off-by Robert Yang <[email protected]>

---
  meta/files/toolchain-shar-extract.sh |    4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)

meta/files/toolchain-shar-extract.sh +2 -2

index 0a33ee8..3a50991 100644
@@ -92,11 +92,11 @@ fi
 
  if [ -e "$target_sdk_dir/environment-setup-@REAL_MULTIMACH_TARGET_SYS@" ]; then
  	echo "The directory \"$target_sdk_dir\" already contains a SDK for this architecture."
-	printf "If you continue, existing files will be overwritten! Proceed[y/N]?"
+	printf "If you continue, existing files will be overwritten! Proceed[y/N]? "
 
  	default_answer="n"
  else
-	printf "You are about to install the SDK to \"$target_sdk_dir\". Proceed[Y/n]?"
+	printf "You are about to install the SDK to \"$target_sdk_dir\". Proceed[Y/n]? "
 
  	default_answer="y"
  fi

Nit pick on logs...

In the short log:
s/add a space in the end/
/Add a space after asking a question/

Quoted 1 line · Robert Yang

For a clear look when input.

This is a sentence fragment. It will read better if you repeat yourself:

Add a space at the end of a line when asking a question. This makes it easier to read.

../Randy

Quoted 31 lines

[YOCTO #8089]

Signed-off-by Robert Yang <[email protected]>

---
  meta/files/toolchain-shar-extract.sh |    4 ++--
  1 file changed, 2 insertions(+), 2 deletions(-)

meta/files/toolchain-shar-extract.sh +0 -0

index 0a33ee8..3a50991 100644
@@ -92,11 +92,11 @@ fi
 
  if [ -e "$target_sdk_dir/environment-setup-@REAL_MULTIMACH_TARGET_SYS@" ]; then
      echo "The directory \"$target_sdk_dir\" already contains a SDK for this

architecture."

meta/files/toolchain-shar-extract.sh +1 -2

-    printf "If you continue, existing files will be overwritten! Proceed[y/N]?"
+    printf "If you continue, existing files will be overwritten! Proceed[y/N]? "
 
      default_answer="n"
  else
-    printf "You are about to install the SDK to \"$target_sdk_dir\".

Proceed[Y/n]?"
+ printf "You are about to install the SDK to \"$target_sdk_dir\".
Proceed[Y/n]? "

default_answer="y"
fi

When configure multilib, "bitbake <image_bb> -c populate_sdk" should install all arch toolchains (for example, 32 and 64bit), but rpm not handle the multilib requires correctly, for example:
lib32-packagegroup-core-standalone-sdk-target requires lib32-libc6, rpm may pull in libc6 rather than lib32-libc6, if we directly use something like IMAGE_INSTALL_append = " lib32-libc6", then it works well, so update TOOLCHAIN_TARGET_TASK to include the RDEPENDS packages will fix the issue.

[YOCTO #8089]

Signed-off-by Robert Yang <[email protected]>

---
 meta/classes/populate_sdk_base.bbclass |   32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)

meta/classes/populate_sdk_base.bbclass +32 -0

index a9e9bd7..c2491ea 100644
@@ -83,8 +83,40 @@ POPULATE_SDK_POST_HOST_COMMAND_append = " write_host_sdk_manifest; "
 fakeroot python do_populate_sdk() {
     from oe.sdk import populate_sdk
     from oe.manifest import create_manifest, Manifest
+    import oe.packagedata
 
     pn = d.getVar('PN', True)
+    # Help rpm to pull in multilib requires since it may not pull in
+    # them correctly, for example,
+    # lib32-packagegroup-core-standalone-sdk-target requires
+    # lib32-libc6, rpm may pull in libc6 rather than lib32-libc6, but
+    # directly use
+    # TOOLCHAIN_TARGET_TASK_append = " lib32-libc6" works well.
+    for pkg in (d.getVar('TOOLCHAIN_TARGET_TASK', True) or "").split():
+        sub_data = oe.packagedata.read_subpkgdata(pkg, d)
+        sub_rdep = sub_data.get("RDEPENDS_" + pkg)
+        if not sub_rdep:
+            continue
+        done = sub_rdep.split()
+        next = done
+        # Find all the rdepends on the dependency chain
+        while next:
+            new = []
+            for sub_pkg in next:
+                sub_data = oe.packagedata.read_subpkgdata(sub_pkg, d)
+                sub_pkg_rdep = sub_data.get("RDEPENDS_" + sub_pkg)
+                if not sub_pkg_rdep:
+                    continue
+                for p in sub_pkg_rdep.split():
+                    if p in done:
+                        continue
+                    if not p.startswith('(') and \
+                            oe.packagedata.has_subpkgdata(p, d):
+                        # It's a new rdep
+                        done.append(p)
+                        new.append(p)
+            next = new
+        d.appendVar('TOOLCHAIN_TARGET_TASK', ' ' + " ".join(done))
     runtime_mapping_rename("TOOLCHAIN_TARGET_TASK", pn, d)
     runtime_mapping_rename("TOOLCHAIN_TARGET_TASK_ATTEMPTONLY", pn, d)
 

Signature · 1 line

1.7.9.5

Quoted 8 lines · Robert Yang

When configure multilib, "bitbake <image_bb> -c populate_sdk" should install all arch toolchains (for example, 32 and 64bit), but rpm not handle the multilib requires correctly, for example:
lib32-packagegroup-core-standalone-sdk-target requires lib32-libc6, rpm may pull in libc6 rather than lib32-libc6, if we directly use something like IMAGE_INSTALL_append = " lib32-libc6", then it works well, so update TOOLCHAIN_TARGET_TASK to include the RDEPENDS packages will fix the issue.

Whilst I understand why you're doing this, its rather sad that as the rpm multilib code is broken, we have to hack the core sdk code like this to work around it.

Could/should we not put this into the rpm sdk backend instead?

I also can't help this is just a bandaid and that we'll likely find other ways this breaks in the future, its not a complete solution to the issue :/.

Cheers,

Richard

Quoted 55 lines

[YOCTO #8089]

Signed-off-by Robert Yang <[email protected]>

---
  meta/classes/populate_sdk_base.bbclass |   32 ++++++++++++++++++++++++++++++++
  1 file changed, 32 insertions(+)

meta/classes/populate_sdk_base.bbclass +32 -1

index a9e9bd7..c2491ea 100644
@@ -83,8 +83,40 @@ POPULATE_SDK_POST_HOST_COMMAND_append = " write_host_sdk_manifest; "
  fakeroot python do_populate_sdk() {
      from oe.sdk import populate_sdk
      from oe.manifest import create_manifest, Manifest
+    import oe.packagedata
 
      pn = d.getVar('PN', True)
+    # Help rpm to pull in multilib requires since it may not pull in
+    # them correctly, for example,
+    # lib32-packagegroup-core-standalone-sdk-target requires
+    # lib32-libc6, rpm may pull in libc6 rather than lib32-libc6, but
+    # directly use
+    # TOOLCHAIN_TARGET_TASK_append = " lib32-libc6" works well.
+    for pkg in (d.getVar('TOOLCHAIN_TARGET_TASK', True) or "").split():
+        sub_data = oe.packagedata.read_subpkgdata(pkg, d)
+        sub_rdep = sub_data.get("RDEPENDS_" + pkg)
+        if not sub_rdep:
+            continue
+        done = sub_rdep.split()
+        next = done
+        # Find all the rdepends on the dependency chain
+        while next:
+            new = []
+            for sub_pkg in next:
+                sub_data = oe.packagedata.read_subpkgdata(sub_pkg, d)
+                sub_pkg_rdep = sub_data.get("RDEPENDS_" + sub_pkg)
+                if not sub_pkg_rdep:
+                    continue
+                for p in sub_pkg_rdep.split():
+                    if p in done:
+                        continue
+                    if not p.startswith('(') and \
+                            oe.packagedata.has_subpkgdata(p, d):
+                        # It's a new rdep
+                        done.append(p)
+                        new.append(p)
+            next = new
+        d.appendVar('TOOLCHAIN_TARGET_TASK', ' ' + " ".join(done))
      runtime_mapping_rename("TOOLCHAIN_TARGET_TASK", pn, d)
      runtime_mapping_rename("TOOLCHAIN_TARGET_TASK_ATTEMPTONLY", pn, d)
 
--

1.7.9.5