[PATCH] utils: Refactor filemode variable conversion to a function

[PATCH] utils: Refactor filemode variable conversion to a function

[Richard Purdie]

We have other places in the code where we need to take filemode/mask
information from a bitbake variable and turn it into a real python
number. Turn this internal code into public API in bb.utils and
add some tests for it.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
---
 bin/bitbake-worker    |  7 ++-----
 lib/bb/tests/utils.py | 11 +++++++++++
 lib/bb/utils.py       | 10 ++++++++++
 3 files changed, 23 insertions(+), 5 deletions(-)

diff --git a/bin/bitbake-worker b/bin/bitbake-worker
index 35fa1ab62be..9f33195176f 100755
--- a/bin/bitbake-worker
+++ b/bin/bitbake-worker
@@ -182,11 +182,8 @@ def fork_off_task(cfg, data, databuilder, workerdata, extraconfigdata, runtask):
     elif workerdata["umask"]:
         umask = workerdata["umask"]
     if umask:
-        # umask might come in as a number or text string..
-        try:
-             umask = int(umask, 8)
-        except TypeError:
-             pass
+        # Convert to a python numeric value as it could be a string
+        bb.utils.to_filemode(umask)
 
     dry_run = cfg.dry_run or runtask['dry_run']
 
diff --git a/lib/bb/tests/utils.py b/lib/bb/tests/utils.py
index 48e61dfcea0..52b7bf85bf4 100644
--- a/lib/bb/tests/utils.py
+++ b/lib/bb/tests/utils.py
@@ -692,3 +692,14 @@ class EnvironmentTests(unittest.TestCase):
         self.assertIn("A", os.environ)
         self.assertEqual(os.environ["A"], "this is A")
         self.assertNotIn("B", os.environ)
+
+class FilemodeTests(unittest.TestCase):
+    def test_filemode_convert(self):
+        self.assertEqual(0o775, bb.utils.to_filemode("0o775"))
+        self.assertEqual(0o775, bb.utils.to_filemode(0o775))
+        self.assertEqual(0o775, bb.utils.to_filemode("775"))
+        with self.assertRaises(ValueError):
+            bb.utils.to_filemode("xyz")
+        with self.assertRaises(ValueError):
+            bb.utils.to_filemode("999")
+
diff --git a/lib/bb/utils.py b/lib/bb/utils.py
index a2806fd360d..01f7cf81f46 100644
--- a/lib/bb/utils.py
+++ b/lib/bb/utils.py
@@ -1211,6 +1211,16 @@ def which(path, item, direction = 0, history = False, executable=False):
         return "", hist
     return ""
 
+def to_filemode(input):
+    """
+    Take a bitbake variable contents defining a file mode and return
+    the proper python representation of the number
+    """
+    # umask might come in as a number or text string..
+    if type(input) is int:
+        return input
+    return int(input, 8)
+
 @contextmanager
 def umask(new_mask):
     """

Re: [bitbake-devel] [PATCH] utils: Refactor filemode variable conversion to a function

[Antonin Godard]

Hi Richard,

On Fri Jun 27, 2025 at 11:16 AM CEST, Richard Purdie via lists.openembedded.org wrote:
> We have other places in the code where we need to take filemode/mask
> information from a bitbake variable and turn it into a real python
> number. Turn this internal code into public API in bb.utils and
> add some tests for it.
>
> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> ---
>  bin/bitbake-worker    |  7 ++-----
>  lib/bb/tests/utils.py | 11 +++++++++++
>  lib/bb/utils.py       | 10 ++++++++++
>  3 files changed, 23 insertions(+), 5 deletions(-)
>
> diff --git a/bin/bitbake-worker b/bin/bitbake-worker
> index 35fa1ab62be..9f33195176f 100755
> --- a/bin/bitbake-worker
> +++ b/bin/bitbake-worker
> @@ -182,11 +182,8 @@ def fork_off_task(cfg, data, databuilder, workerdata, extraconfigdata, runtask):
>      elif workerdata["umask"]:
>          umask = workerdata["umask"]
>      if umask:
> -        # umask might come in as a number or text string..
> -        try:
> -             umask = int(umask, 8)
> -        except TypeError:
> -             pass
> +        # Convert to a python numeric value as it could be a string
> +        bb.utils.to_filemode(umask)
>  
>      dry_run = cfg.dry_run or runtask['dry_run']
>  
> diff --git a/lib/bb/tests/utils.py b/lib/bb/tests/utils.py
> index 48e61dfcea0..52b7bf85bf4 100644
> --- a/lib/bb/tests/utils.py
> +++ b/lib/bb/tests/utils.py
> @@ -692,3 +692,14 @@ class EnvironmentTests(unittest.TestCase):
>          self.assertIn("A", os.environ)
>          self.assertEqual(os.environ["A"], "this is A")
>          self.assertNotIn("B", os.environ)
> +
> +class FilemodeTests(unittest.TestCase):
> +    def test_filemode_convert(self):
> +        self.assertEqual(0o775, bb.utils.to_filemode("0o775"))
> +        self.assertEqual(0o775, bb.utils.to_filemode(0o775))
> +        self.assertEqual(0o775, bb.utils.to_filemode("775"))
> +        with self.assertRaises(ValueError):
> +            bb.utils.to_filemode("xyz")
> +        with self.assertRaises(ValueError):
> +            bb.utils.to_filemode("999")
> +
> diff --git a/lib/bb/utils.py b/lib/bb/utils.py
> index a2806fd360d..01f7cf81f46 100644
> --- a/lib/bb/utils.py
> +++ b/lib/bb/utils.py
> @@ -1211,6 +1211,16 @@ def which(path, item, direction = 0, history = False, executable=False):
>          return "", hist
>      return ""
>  
> +def to_filemode(input):
> +    """
> +    Take a bitbake variable contents defining a file mode and return
> +    the proper python representation of the number
> +    """

It could be nice to keep the same format as the other functions in this file:
explicit the arguments and the return value.

> +    # umask might come in as a number or text string..
> +    if type(input) is int:
> +        return input
> +    return int(input, 8)
> +
>  @contextmanager
>  def umask(new_mask):
>      """

I think this triggered a wall of errors of the autobuilder:
https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/1910

Locally I also reproduced the issue:

ERROR: Task (/data/yoctoproject/ws/repos/openembedded-core/meta/recipes-bsp/efivar/efivar_39.bb:do_recipe_qa) failed with exit code '1'

If I drop this commit (and the two others on OE-Core that use to_filemode),
then re-run it becomes successful again. Cleaning the sstate of the recipe
then, and re-introducing your commit only, triggers the error. I don't have any
log to show because I think the issue happens before do_recipe_qa is run.

Antonin

-- 
Antonin Godard, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com

Re: [bitbake-devel] [PATCH] utils: Refactor filemode variable conversion to a function

[Richard Purdie]

On Mon, 2025-06-30 at 09:42 +0200, Antonin Godard wrote:
> Hi Richard,
> 
> On Fri Jun 27, 2025 at 11:16 AM CEST, Richard Purdie via lists.openembedded.org wrote:
> > We have other places in the code where we need to take filemode/mask
> > information from a bitbake variable and turn it into a real python
> > number. Turn this internal code into public API in bb.utils and
> > add some tests for it.
> > 
> > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> > ---
> >  bin/bitbake-worker    |  7 ++-----
> >  lib/bb/tests/utils.py | 11 +++++++++++
> >  lib/bb/utils.py       | 10 ++++++++++
> >  3 files changed, 23 insertions(+), 5 deletions(-)
> > 
> > diff --git a/bin/bitbake-worker b/bin/bitbake-worker
> > index 35fa1ab62be..9f33195176f 100755
> > --- a/bin/bitbake-worker
> > +++ b/bin/bitbake-worker
> > @@ -182,11 +182,8 @@ def fork_off_task(cfg, data, databuilder, workerdata, extraconfigdata, runtask):
> >      elif workerdata["umask"]:
> >          umask = workerdata["umask"]
> >      if umask:
> > -        # umask might come in as a number or text string..
> > -        try:
> > -             umask = int(umask, 8)
> > -        except TypeError:
> > -             pass
> > +        # Convert to a python numeric value as it could be a string
> > +        bb.utils.to_filemode(umask)
> >  
> >      dry_run = cfg.dry_run or runtask['dry_run']
> >  
> > diff --git a/lib/bb/tests/utils.py b/lib/bb/tests/utils.py
> > index 48e61dfcea0..52b7bf85bf4 100644
> > --- a/lib/bb/tests/utils.py
> > +++ b/lib/bb/tests/utils.py
> > @@ -692,3 +692,14 @@ class EnvironmentTests(unittest.TestCase):
> >          self.assertIn("A", os.environ)
> >          self.assertEqual(os.environ["A"], "this is A")
> >          self.assertNotIn("B", os.environ)
> > +
> > +class FilemodeTests(unittest.TestCase):
> > +    def test_filemode_convert(self):
> > +        self.assertEqual(0o775, bb.utils.to_filemode("0o775"))
> > +        self.assertEqual(0o775, bb.utils.to_filemode(0o775))
> > +        self.assertEqual(0o775, bb.utils.to_filemode("775"))
> > +        with self.assertRaises(ValueError):
> > +            bb.utils.to_filemode("xyz")
> > +        with self.assertRaises(ValueError):
> > +            bb.utils.to_filemode("999")
> > +
> > diff --git a/lib/bb/utils.py b/lib/bb/utils.py
> > index a2806fd360d..01f7cf81f46 100644
> > --- a/lib/bb/utils.py
> > +++ b/lib/bb/utils.py
> > @@ -1211,6 +1211,16 @@ def which(path, item, direction = 0, history = False, executable=False):
> >          return "", hist
> >      return ""
> >  
> > +def to_filemode(input):
> > +    """
> > +    Take a bitbake variable contents defining a file mode and return
> > +    the proper python representation of the number
> > +    """
> 
> It could be nice to keep the same format as the other functions in this file:
> explicit the arguments and the return value.
> 
> > +    # umask might come in as a number or text string..
> > +    if type(input) is int:
> > +        return input
> > +    return int(input, 8)
> > +
> >  @contextmanager
> >  def umask(new_mask):
> >      """
> 
> I think this triggered a wall of errors of the autobuilder:
> https://autobuilder.yoctoproject.org/valkyrie/#/builders/29/builds/1910
> 
> Locally I also reproduced the issue:
> 
> ERROR: Task (/data/yoctoproject/ws/repos/openembedded-core/meta/recipes-bsp/efivar/efivar_39.bb:do_recipe_qa) failed with exit code '1'
> 
> If I drop this commit (and the two others on OE-Core that use to_filemode),
> then re-run it becomes successful again. Cleaning the sstate of the recipe
> then, and re-introducing your commit only, triggers the error. I don't have any
> log to show because I think the issue happens before do_recipe_qa is run.

Thanks, I've tweaked the docstring and fixed the issue in v2.

Cheers,

Richard