Skip to contentSkip to conversation

cooker.py: Remove explicit build targets from the ignored list

PATCH In review Mark Hatle 4 participants 5 messages May 29, 2013 – Jun 4, 2013

lore
Share & tools
lore

Mark Hatle

author
#0 lore raw

PATCH 1/1 cooker.py: Remove explicit build targets from the ignored list

From: Peter Seebach <[email protected]>

If something is in ASSUMED_PROVIDED, and a user tries to build it explicitly, Nothing Happens. Bitbake just says it ran 0 out of 0 tasks. No diagnostics or explanations are offered.

In practice, the right thing is probably to assume that explicit targets are intentional. So far as I can tell, cooker.buildTargets is called only from the command line or from command.py, and both cases seem to be explicit user actions.

Rename 'status' to 'recipecache' to match recent changes.

---
 lib/bb/cooker.py | 5 +++++
 1 file changed, 5 insertions(+)
lib/bb/cooker.py +5 -0
index 2ae3e9d..8f0e90c 100644
@@ -1067,6 +1067,11 @@ class BBCooker:
         universe = ('universe' in targets)
         targets = self.checkPackages(targets)
 
+        # Explicitly disignore things that have been requested.
+        for target in targets:
+            if target in self.recipecache.ignored_dependencies:
+                self.recipecache.ignored_dependencies.remove(target)
+
         def buildTargetsIdle(server, rq, abort):
             if abort or self.state == state.stop:
                 rq.finish_runqueue(True)
Signature · 1 line
1.8.1.2.545.g2f19ada
#1 lore raw
Quoted 27 lines · Mark Hatle

From: Peter Seebach <[email protected]>

If something is in ASSUMED_PROVIDED, and a user tries to build it explicitly, Nothing Happens. Bitbake just says it ran 0 out of 0 tasks. No diagnostics or explanations are offered.

In practice, the right thing is probably to assume that explicit targets are intentional. So far as I can tell, cooker.buildTargets is called only from the command line or from command.py, and both cases seem to be explicit user actions.

Rename 'status' to 'recipecache' to match recent changes.

---
 lib/bb/cooker.py | 5 +++++
 1 file changed, 5 insertions(+)
lib/bb/cooker.py +0 -0
index 2ae3e9d..8f0e90c 100644
@@ -1067,6 +1067,11 @@ class BBCooker:
         universe = ('universe' in targets)

lib/bb/cooker.py class BBCooker: · line 1077

Quoted 10 lines
lib/bb/cooker.py +5 -0
         targets = self.checkPackages(targets)
 
+        # Explicitly disignore things that have been requested.
+        for target in targets:
+            if target in self.recipecache.ignored_dependencies:
+                self.recipecache.ignored_dependencies.remove(target)
+
         def buildTargetsIdle(server, rq, abort):
             if abort or self.state == state.stop:
                 rq.finish_runqueue(True)

I'm not sure this is the right thing to do, the reason being determinism. I don't expect different results for X in the different cases:

bitbake X Y

bitbake X

(assuming Y is ASSUME_PROVIDED)

For example, X will rebuild between these two commands since the sstate and task checksums will be different between the two. This really needs more thought to make a consistent user experience...

Cheers,

Richard

#2 lore raw
Quoted 48 lines · On Friday 31 May 2013 08:29:43 Richard Purdie wrote:
Quoted 39 lines · Mark Hatle

From: Peter Seebach <[email protected]>

If something is in ASSUMED_PROVIDED, and a user tries to build it explicitly, Nothing Happens. Bitbake just says it ran 0 out of 0 tasks. No diagnostics or explanations are offered.

In practice, the right thing is probably to assume that explicit targets are intentional. So far as I can tell, cooker.buildTargets is called only from the command line or from command.py, and both cases seem to be explicit user actions.

Rename 'status' to 'recipecache' to match recent changes.

---

 lib/bb/cooker.py | 5 +++++
 1 file changed, 5 insertions(+)
lib/bb/cooker.py +5 -0
index 2ae3e9d..8f0e90c 100644
@@ -1067,6 +1067,11 @@ class BBCooker:
         universe = ('universe' in targets)
         targets = self.checkPackages(targets)
 
+        # Explicitly disignore things that have been requested.
+        for target in targets:
+            if target in self.recipecache.ignored_dependencies:
+                self.recipecache.ignored_dependencies.remove(target)
+
 
         def buildTargetsIdle(server, rq, abort):
             if abort or self.state == state.stop:
Quoted 1 line

rq.finish_runqueue(True)

I'm not sure this is the right thing to do, the reason being determinism. I don't expect different results for X in the different cases:

bitbake X Y

Quoted 7 lines

bitbake X

(assuming Y is ASSUME_PROVIDED)

For example, X will rebuild between these two commands since the sstate and task checksums will be different between the two. This really needs more thought to make a consistent user experience...

It seems to me the most consistent behaviour would be to just report a warning that Y is in ASSUME_PROVIDED and carry on building.

Cheers,
Paul

Signature · 2 lines
Paul Eggleton
Intel Open Source Technology Centre
#3 lore raw
In reply to Richard Purdie
Quoted 6 lines · On Fri, 31 May 2013 08:29:43 +0100 Richard Purdie…

I'm not sure this is the right thing to do, the reason being determinism. I don't expect different results for X in the different cases:

bitbake X Y

Quoted 7 lines

bitbake X

(assuming Y is ASSUME_PROVIDED)

For example, X will rebuild between these two commands since the sstate and task checksums will be different between the two. This really needs more thought to make a consistent user experience...

Hmm. I don't *normally* expect different results, but might if the first one produced a diagnostic: "Warning: Y is in ASSUME_PROVIDED. Rebuilding Y will produce different versions of anything which depends on Y."

Paul's suggestion that we just print a diagnostic for anything which is in both targets and ASSUME_PROVIDED might well do, too. Although it might be nice to have a reasonably straight forward command-line way to express "no, really, I want you to build Y" which doesn't require config file action.

-s

Signature · 1 line
Listen, get this.  Nobody with a good compiler needs to be justified.
#4 lore raw
Quoted 17 lines · Peter Seebach
Quoted 6 lines · On Fri, 31 May 2013 08:29:43 +0100 Richard Purdie…

I'm not sure this is the right thing to do, the reason being determinism. I don't expect different results for X in the different cases:

bitbake X Y

Quoted 7 lines

bitbake X

(assuming Y is ASSUME_PROVIDED)

For example, X will rebuild between these two commands since the sstate and task checksums will be different between the two. This really needs more thought to make a consistent user experience...

Quoted 6 lines

Hmm. I don't *normally* expect different results, but might if the first one produced a diagnostic: "Warning: Y is in ASSUME_PROVIDED. Rebuilding Y will produce different versions of anything which depends on Y."

Paul's suggestion that we just print a diagnostic for anything which is in both targets and ASSUME_PROVIDED might well do, too.

I'll certainly take a patch for that.

Quoted 3 lines

Although it might be nice
to have a reasonably straight forward command-line way to express "no, really, I want you to build Y" which doesn't require config file action.

Agreed, however we need to get the user interaction right here. Currently the user doesn't expect for example that the configuration files change from under them. In future this may well be something we do (along with informing the user what happened). Users tend not to read information though :/.

Cheers,

Richard