bitbake.git
8 years agohob: allow configuring default machine using HOB_MACHINE
Paul Eggleton [Fri, 21 Sep 2012 14:47:24 +0000 (15:47 +0100)]
hob: allow configuring default machine using HOB_MACHINE

Allow specifying HOB_MACHINE in local.conf to set the initially
selected machine. With this set, Hob will select the specified machine
and then jump straight into parsing recipes. If you do wish to change
the selected machine with HOB_MACHINE set you still can - you just need
to stop the parsing process first.

Fixes [YOCTO #3148].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: don't show error dialog for errors during building
Paul Eggleton [Fri, 21 Sep 2012 14:47:23 +0000 (15:47 +0100)]
hob: don't show error dialog for errors during building

During building we already report errors in a special tab and
indicate when the build has failed; bringing up a dialog was a
regression introduced in bitbake revision
5bab81b124087d63d6eb62a861e1241714fcd483.

Fixes [YOCTO #3151].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: The title of the packages screen depends on the screen you arrive from
Constantin Musca [Fri, 21 Sep 2012 10:28:43 +0000 (13:28 +0300)]
hob: The title of the packages screen depends on the screen you arrive from

- If you arrive to the packages screen from the recipes screen, the title
should say: 'Step 2 of 2: Edit packages'
- If you arrive to the packages screen from the image details screen, the
title should say: 'Edit packages'
- The title of the recipes screen should say 'Step 1 of 2: Edit recipes'

[YOCTO #2982]

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: Don't set busy cursor on the root window, just our window
Ross Burton [Tue, 18 Sep 2012 14:55:24 +0000 (15:55 +0100)]
hob: Don't set busy cursor on the root window, just our window

[ YOCTO #3127 ]

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: add a top bar when building process is stopped
Cristiana Voicu [Tue, 18 Sep 2012 15:15:55 +0000 (18:15 +0300)]
hob: add a top bar when building process is stopped

When a build was stopped, it wasn't obvious what to do next.
Now, on the page it appers a top bar with 3 buttons: "edit image",
"open log", "build new image"

[YOCTO #2537]
Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocooker: Clarify package/recipe wording
Trevor Woerner [Wed, 19 Sep 2012 03:00:33 +0000 (23:00 -0400)]
cooker: Clarify package/recipe wording

When the '-s' option is run, change the heading above the list of recipes
to say "Recipe Name" instead of "Package Name".

Signed-off-by: Trevor Woerner <twoerner@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/packageselectionpage: "Cancel" button returns to "Image configuration" screen
Cristiana Voicu [Wed, 19 Sep 2012 07:36:57 +0000 (10:36 +0300)]
hob/packageselectionpage: "Cancel" button returns to "Image configuration" screen

Once package building completes, you can customise the list of
packages that will go into the final image. Once you have made
the changes you need, you can either build your image, or you
can exit the process and go back to the 'Image configuration'
screen by selecting 'Cancel'.

[YOCTO #3105]
Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: report event handler failures
Paul Eggleton [Wed, 19 Sep 2012 10:25:42 +0000 (11:25 +0100)]
hob: report event handler failures

If an event handler failed we were not recieving an error message -
parsing just "froze" at 99%. This is because we were expecting a
CommandFailure event and this never happened in the case of
RequestPackageInfo which is where the failure was occurring.

This also required tweaking the error formatting slightly, taking the
return value of the format function rather than the message property
since the latter only seems to contain the first line without the
traceback in the case of event handler failure. Other error cases were
tested and their message formatting is unaffected by this change.

Final part of the fix for [YOCTO #2651].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/settings: Remove stray "distro" combobox from simple settings
Bogdan Marinescu [Thu, 20 Sep 2012 10:27:56 +0000 (13:27 +0300)]
hob/settings: Remove stray "distro" combobox from simple settings

The "distro" combobox was moved to advanced settings, but it
was also present in simple settings. This patch removed it
from simple settings.

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobuild/siggen: Add support for stamp 'clean' masks
Richard Purdie [Tue, 18 Sep 2012 10:32:04 +0000 (11:32 +0100)]
build/siggen: Add support for stamp 'clean' masks

Currently when we execute a task, we don't remove other potentially stale
stamps. This can mean if you switch between two different versions of a
recipe without a clean, the build can get very confused.

This patch adds in functionality to allow a wildcard expression of stamp
files to be removed when creating a new stamp file. This patch adds in
the core of the code to enable this but it also requires metadata support
to enable it.

When writing this improvement I went through several different options but
this was the only way I could find to allow things like noexec tasks to
function correctly (where stamps need to be created without the data store).

[YOCTO #2961]

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2/cvs: Fix parameter spacing
Richard Purdie [Thu, 20 Sep 2012 10:34:18 +0000 (11:34 +0100)]
fetch2/cvs: Fix parameter spacing

Add in misssing space between the parameters. Reported by
Jate Sujjavanich <Jate.Sujjavanich@myfuelmaster.com>.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2/cvs: Clean up various data store references
Richard Purdie [Thu, 20 Sep 2012 10:32:02 +0000 (11:32 +0100)]
fetch2/cvs: Clean up various data store references

The code in the CVS fetcher is elderly and there are simpler ways of
using the data store. This updates to use the modern APIs.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2/cvs: Fix localdata variable reference
Richard Purdie [Thu, 20 Sep 2012 10:25:19 +0000 (11:25 +0100)]
fetch2/cvs: Fix localdata variable reference

The localdata variable was removed, fix up a lost reference to this.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: Fix settings dialog issues
Bogdan Marinescu [Fri, 14 Sep 2012 14:17:13 +0000 (17:17 +0300)]
hob: Fix settings dialog issues

Fix some issues with the settings dialog that were introduced as
a result of merging the fix for #2162.

[YOCTO #3117]

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: rename 'View log' to 'Open log'
Constantin Musca [Fri, 14 Sep 2012 09:58:53 +0000 (12:58 +0300)]
hob: rename 'View log' to 'Open log'

Rename all the 'View log' buttons to 'Open log' for
consistency.

[YOCTO #3045]

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocompat, event: use OrderedDict from py2.7 for the event handlers
Christopher Larson [Mon, 10 Sep 2012 22:23:19 +0000 (22:23 +0000)]
compat, event: use OrderedDict from py2.7 for the event handlers

This ensures that our event handlers get run in registration order, making the
behavior more deterministic. I pulled in the python2.7 OrderedDict to avoid
essentially reimplementing a version of it ourselves, figuring we can drop it
when we bump our required python version next.

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoImplement 'settings' dialog as designed
Valentin Popa [Thu, 6 Sep 2012 07:58:32 +0000 (10:58 +0300)]
Implement 'settings' dialog as designed

[YOCTO #2162]

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocooker: Fix priority for virtual recipes
Richard Purdie [Thu, 13 Sep 2012 14:25:51 +0000 (15:25 +0100)]
cooker: Fix priority for virtual recipes

When making filename comparisons for recipes for priority calculations,
we need to split off any virtual prefix. Without this, BBCLASSEXTEND
version of recipes don't follow the priority settings they should.

[YOCTO #2933]

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/builddetailspage: Add tooltips to the build failed notification
Constantin Musca [Thu, 13 Sep 2012 11:33:34 +0000 (14:33 +0300)]
hob/builddetailspage: Add tooltips to the build failed notification

[YOCTO #3046]

Signed-off-by: Constantin Musca <constantinx.musca@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolib/bb/event: improve handling of event queue on exit
Paul Eggleton [Tue, 11 Sep 2012 09:37:22 +0000 (10:37 +0100)]
lib/bb/event: improve handling of event queue on exit

If BitBake exits before a UI handler (server) has been registered, we
print the event queue; if there are any errors or other non-debug
messages just print these and suppress the rest of the message queue.

This improves the output when sanity check failures occur with OE-Core
by avoiding printing a long stream of uninformative debug messages.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: format error messages properly
Paul Eggleton [Tue, 11 Sep 2012 11:00:05 +0000 (12:00 +0100)]
hob: format error messages properly

Error messages that use arguments need to be formatted properly, or we
don't get the full message. Use a formatter to do this when an error
occurs.

Partial fix for [YOCTO #2983].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/imagedetailspage: change "FileCreated" label to "Files created"
Cristiana Voicu [Tue, 11 Sep 2012 07:40:06 +0000 (10:40 +0300)]
hob/imagedetailspage: change "FileCreated" label to "Files created"

[YOCTO #2998]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoFixed hob proxy tab tooltip text per review suggestions [YOCTO #2499]
Jessica Zhang [Tue, 11 Sep 2012 23:55:52 +0000 (16:55 -0700)]
Fixed hob proxy tab tooltip text per review suggestions [YOCTO #2499]

Signed-off-by: Jessica Zhang <jessica.zhang@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: sort base image drop-down list
Paul Eggleton [Tue, 11 Sep 2012 16:52:01 +0000 (17:52 +0100)]
hob: sort base image drop-down list

Sort the list of base images to make it easier to find a specific image
in the list. Note that "Create your own image" still remains the last
item in the list.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolib/bb/command.py: ensure setVariable only sets values as strings
Paul Eggleton [Tue, 11 Sep 2012 16:52:00 +0000 (17:52 +0100)]
lib/bb/command.py: ensure setVariable only sets values as strings

This is the interface Hob uses to set variable values in many instances,
and at the moment it is possible that some of the values it passes are
not strings. If a non-string value gets into the datastore it can
trigger exceptions during parsing when we attempt to expand the variable
and substitute in the non-string value.

This fixes using the meta-ti layer within Hob - it currently has a
reference to BB_NUMBER_THREADS within a shell function and since this
is a variable that Hob was setting from its configuration as an integer,
due to the above this was triggering an ExpansionError.

Fixes [YOCTO #2875].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: don't reorder layers list
Paul Eggleton [Tue, 11 Sep 2012 16:51:59 +0000 (17:51 +0100)]
hob: don't reorder layers list

We cannot reorder this list - it must stay in the order shown in the
dialog (which may in future be configurable by the user).

Fixes [YOCTO #2649].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/packageselectionpage: restore selected packages
Cristiana Voicu [Mon, 10 Sep 2012 12:49:57 +0000 (15:49 +0300)]
hob/packageselectionpage: restore selected packages

After "Cancel" action, selected packages are restored to default.

[YOCTO #2984]
Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: rename task -> packagegroup in recipe selection
Paul Eggleton [Mon, 10 Sep 2012 17:26:46 +0000 (18:26 +0100)]
hob: rename task -> packagegroup in recipe selection

This changes the filtering to use the inheritance of
packagegroup.bbclass to determine if a recipe is a package group.

Also makes the tab tooltip text generic; these recipes could come from
any enabled layer, not just the default ones.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: ensure error message text is properly escaped
Paul Eggleton [Mon, 10 Sep 2012 17:11:17 +0000 (18:11 +0100)]
hob: ensure error message text is properly escaped

Our poor implementation of markup escaping was causing invalid
markup, leading to the error dialog being blank. Use the glib markup
escaping function provided by PyGTK+ to do this properly and avoid the
blank error dialogs.

Partial fix for [YOCTO #2983].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: use correct semantics for dealing with pkgdata
Paul Eggleton [Mon, 10 Sep 2012 14:36:29 +0000 (15:36 +0100)]
hob: use correct semantics for dealing with pkgdata

Some of these values may or may not be overridden on a per-package
basis, so handle them accordingly.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agotinfoil: Add file inadvertently not committed
Richard Purdie [Mon, 10 Sep 2012 20:50:34 +0000 (21:50 +0100)]
tinfoil: Add file inadvertently not committed

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocooker: fix handling of exceptions during exception handling
Paul Eggleton [Fri, 7 Sep 2012 15:22:54 +0000 (16:22 +0100)]
cooker: fix handling of exceptions during exception handling

If an exception occurs during handling another exception we were
getting a useless traceback such as the following, after which
BitBake froze:

ERROR: Command execution failed: Traceback (most recent call last):
  File "/home/user/poky/poky/bitbake/lib/bb/command.py", line 84, in runAsyncCommand
    self.cooker.updateCache()
  File "/home/user/poky/poky/bitbake/lib/bb/cooker.py", line 1207, in updateCache
    if not self.parser.parse_next():
  File "/home/user/poky/poky/bitbake/lib/bb/cooker.py", line 1694, in parse_next
    logger.error('Unable to parse %s', value.recipe,
AttributeError: 'exceptions.TypeError' object has no attribute 'recipe'

Fix this to print an actual traceback of the exception and exit
gracefully (well, as gracefully as possible under the circumstances).

The general fix for [YOCTO #2977].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2: fix malformed URL causing a useless traceback
Paul Eggleton [Fri, 7 Sep 2012 15:22:53 +0000 (16:22 +0100)]
fetch2: fix malformed URL causing a useless traceback

The implementation of NoMethodError and MalformedUrl was broken - if you
just set self.args in an exception class to a string it treats it as a
list and then fails later on with a TypeError due to the number of
arguments not matching up.

This nasty exception during exception handling was breaking the normal
exception flow (fixed separately), which meant that if you had a
malformed URL or invalid protocol in SRC_URI you would get the
following:

ERROR: Command execution failed: Traceback (most recent call last):
  File "/home/user/poky/poky/bitbake/lib/bb/command.py", line 84, in runAsyncCommand
    self.cooker.updateCache()
  File "/home/user/poky/poky/bitbake/lib/bb/cooker.py", line 1207, in updateCache
    if not self.parser.parse_next():
  File "/home/user/poky/poky/bitbake/lib/bb/cooker.py", line 1694, in parse_next
    logger.error('Unable to parse %s', value.recipe,
AttributeError: 'exceptions.TypeError' object has no attribute 'recipe'

A specific fix for [YOCTO #2977].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/imageconfigurationpage: remove or_label reference
Cristiana Voicu [Mon, 10 Sep 2012 08:49:27 +0000 (11:49 +0300)]
hob/imageconfigurationpage: remove or_label reference

When or_label was removed, I forgot to remove also the references to it.

[YOCTO #3010]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2: replace double slashes in paths in encodeurl()
Paul Eggleton [Fri, 7 Sep 2012 13:34:35 +0000 (14:34 +0100)]
fetch2: replace double slashes in paths in encodeurl()

This ensures that if all a MIRRORS entry does is add a slash, this does
not result in a circular loop.

Fixes [YOCTO #3073].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: print message when DISPLAY is not set
Cristiana Voicu [Fri, 7 Sep 2012 12:00:11 +0000 (15:00 +0300)]
hob: print message when DISPLAY is not set

If DISPLAY wasn't set, launching hob has printed a traceback
difficult to understand. Now, the exception is caught and it
shows a human message.

[YOCTO #2596]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoHob: change view of 'recipes' and 'packages' tables as ui design
Liming An [Wed, 5 Sep 2012 11:13:48 +0000 (19:13 +0800)]
Hob: change view of 'recipes' and 'packages' tables as ui design

changed the order of task tables,
cancel the 'description' column,
add the binb total number indicator, and so on

[YOCTO 2195]

Signed-off-by: Liming An <limingx.l.an@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobitbake-diffsigs: allow specifying task & follow deps recursively
Paul Eggleton [Mon, 27 Aug 2012 20:44:35 +0000 (21:44 +0100)]
bitbake-diffsigs: allow specifying task & follow deps recursively

Add the ability to compare the two most recent runs of a specified task,
and follow dependent hash changes recursively. This enables you to trace
back and find exactly why a task was re-run after the fact.

Note that this relies on the metadata providing a function, hooked in
as bb.siggen.find_siginfo, which allows searching in the appropriate
places to find signature data files.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolib/bb/siggen.py: make signature dump/compare functions return a list
Paul Eggleton [Mon, 27 Aug 2012 20:44:34 +0000 (21:44 +0100)]
lib/bb/siggen.py: make signature dump/compare functions return a list

These functions become a little bit more reusable if they return a list
containing the output rather than just printing it.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolib/bb/siggen.py: insert a colon between class and recipe name
Paul Eggleton [Mon, 27 Aug 2012 20:44:33 +0000 (21:44 +0100)]
lib/bb/siggen.py: insert a colon between class and recipe name

before:
  virtual:nativeautomake_1.12.1.bb.do_compile
after:
  virtual:native:automake_1.12.1.bb.do_compile

This separation ensures that the key is readable, and if necessary,
parsable.

Unfortunately this invalidates previous native sstate signatures with
OE-Core - not much that can be done about that; however that occurs
frequently during the development cycle so it's par for the course.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agolib/bb/siggen: replace tabs with spaces
Paul Eggleton [Mon, 27 Aug 2012 20:44:32 +0000 (21:44 +0100)]
lib/bb/siggen: replace tabs with spaces

We had one section of the code mixing tabs with spaces, which is
particularly undesirable with python code.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agotinfoil: create simple interface for bitbake-based utilities
Paul Eggleton [Mon, 27 Aug 2012 20:44:31 +0000 (21:44 +0100)]
tinfoil: create simple interface for bitbake-based utilities

The code to initialise BitBake within bitbake-layers should be useful
for other utilities that need to query configuration or recipe
information, so refactor it out into its own class, "Tinfoil" (to
continue with our cooking metaphor).

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocrumbs/hig: Save toolchain in settings
Bogdan Marinescu [Wed, 5 Sep 2012 10:11:25 +0000 (13:11 +0300)]
crumbs/hig: Save toolchain in settings

Toolchain was not saved in the settings dialog ("Output" tab).

[YOCTO #2695]

Signed-off-by: Bogdan Marinescu <bogdan.a.marinescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: The 'run image' and 'deploy image' dialogs text and alignment corrections
Cristiana Voicu [Wed, 5 Sep 2012 07:50:33 +0000 (10:50 +0300)]
hob: The 'run image' and 'deploy image' dialogs text and alignment corrections

-changed the text shown by both dialogs text
-make small tweaks to alignment

[YOCTO #2999]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/imagedetailspage: "Image ready" icon appers only after the image was generated
Cristiana Voicu [Mon, 3 Sep 2012 11:22:04 +0000 (14:22 +0300)]
hob/imagedetailspage: "Image ready" icon appers only after the image was generated

Now, "Your image is ready" icon doesn't appear when you come back on Imagedetails
page. It appears only after the image was generated.

[YOCTO #2984]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohobwidget: Button theme is taken from host, fix
Ioana Grigoropol [Mon, 3 Sep 2012 09:51:06 +0000 (12:51 +0300)]
hobwidget: Button theme is taken from host, fix

- All buttons in the interface inherit a BaseHobButton that
  use the gtk settings for buttons from the host;
- Removed 'or' label between actions on image details page

[Yocto #3011]

Signed-off-by: Ioana Grigoropol <ioanax.grigoropol@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2: unpack rpm, ipk and deb binary package
Robert Yang [Sun, 2 Sep 2012 04:08:51 +0000 (12:08 +0800)]
fetch2: unpack rpm, ipk and deb binary package

* Unpack the ".rpm" binary package (only .src.rpm in the past)

* Unpack the .deb and .ipk binary package, their unpack commands are the same.

* This is useful for binary package recipe.

[YOCTO #1592]

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/recipe&packageselectionpage: Change "Back" button to "Cancel" button
Cristiana Voicu [Fri, 31 Aug 2012 14:10:11 +0000 (17:10 +0300)]
hob/recipe&packageselectionpage: Change "Back" button to "Cancel" button

"Back" button placed on RecipeSelectionPage and PackageSelectionPage
was changed to "Cancel" button to avoid any confusion.
Also, it was placed next to the other buttons on the page.

[YOCTO #3012]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/imageconfigurationpage: Set secondary action for "Edit image" button
Cristiana Voicu [Fri, 31 Aug 2012 11:37:58 +0000 (14:37 +0300)]
hob/imageconfigurationpage: Set secondary action for "Edit image" button

The image configuration screen should have only one primary action.
"Edit image" button has now secondary action, and also I have removed
"or" label.

[YOCTO #3010]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/packageselectionpage: Add tooltips to 'Included' and 'All packages' tab and ...
Cristiana Voicu [Fri, 31 Aug 2012 08:48:23 +0000 (11:48 +0300)]
hob/packageselectionpage: Add tooltips to 'Included' and 'All packages' tab and 'Search' field

For the 'Included' tab: "The packages currently included for your image"
For the 'All packages' tab: "All packages that have been built"
For the 'Search' field: "Enter a package name to find it"

[YOCTO #2322]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob/packageselectionpage: Correctly restore previously selected packages
Cristiana Voicu [Fri, 31 Aug 2012 07:00:47 +0000 (10:00 +0300)]
hob/packageselectionpage: Correctly restore previously selected packages

"Back" button from "Packageselection" page now restores correctly previously
selected packages list. Till now "Back" button was implemented just to switch
pages, not to cancel changes you have made to packages list.

[YOCTO #2984]

Signed-off-by: Cristiana Voicu <cristiana.voicu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob2: remove class hcc
Kang Kai [Tue, 28 Aug 2012 02:47:20 +0000 (10:47 +0800)]
hob2: remove class hcc

Because class hcc is useless, remove it.

Signed-off-by: Kang Kai <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob2: remove the hard-coded images map
Kang Kai [Tue, 28 Aug 2012 02:47:19 +0000 (10:47 +0800)]
hob2: remove the hard-coded images map

[Yocto #2795]

When a new image type added, the hob will crash because the new type is
not in the hard-coded image dictionary.

For most of the image types, they are same with the image file's
extension name. So use variable "IMAGE_EXTENSION_difftype" to map the
image type which is diff with the image file extension name, such as
type "live". And the variable(s) will be set in image_types.bbclass.

Signed-off-by: Kang Kai <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agotaskdata: Don't add dependencies on tasks that don't exist
Richard Purdie [Tue, 28 Aug 2012 15:11:34 +0000 (08:11 -0700)]
taskdata: Don't add dependencies on tasks that don't exist

"bitbake meta-toolchain" with qemu image testing enabled causes problems
since it adds a task after do_rootfs which doesn't exist in this case.

We should simply ignore these extra dependencies rather than adding
them in which is what this patch does (adding a debug message when this
happens).

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobin/bitbake: Clarify wording in help output for -s
Trevor Woerner [Sat, 25 Aug 2012 23:26:45 +0000 (19:26 -0400)]
bin/bitbake: Clarify wording in help output for -s

The '-s' option shows (input) recipes, not (built/output) packages. Clarify
the help wording for this option so it is consistent with how bitbake defines
recipes and packages.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoast: Store anonymous python function contents in the datstore
Richard Purdie [Wed, 22 Aug 2012 19:01:55 +0000 (20:01 +0100)]
ast: Store anonymous python function contents in the datstore

This is useful if we need to disable part of one during a backtrace
for debugging purposes.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoast: Extract text variable in PythonMethodNode
Richard Purdie [Wed, 22 Aug 2012 12:54:47 +0000 (13:54 +0100)]
ast: Extract text variable in PythonMethodNode

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoast: Rename PythonMethodNode define variable to modulename
Richard Purdie [Wed, 22 Aug 2012 12:54:01 +0000 (13:54 +0100)]
ast: Rename PythonMethodNode define variable to modulename

It was hard for me to understand what the define variable was, modulename
is hopefully a bit better.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agomethodpool: Clean up the parsed module list handling to be slightly less insane
Richard Purdie [Wed, 22 Aug 2012 12:53:11 +0000 (13:53 +0100)]
methodpool: Clean up the parsed module list handling to be slightly less insane

This removes some dubious functions and replaces them with a simpler, cleaner
API which better describes what the code is doing. Unused code/variables are
removed and comments tweaked.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agomethodpool: Remove unused check_insert_method function
Richard Purdie [Wed, 22 Aug 2012 12:52:24 +0000 (13:52 +0100)]
methodpool: Remove unused check_insert_method function

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocooker: Add explict handling of BBHandledException for parsing failures
Richard Purdie [Wed, 22 Aug 2012 21:40:14 +0000 (22:40 +0100)]
cooker: Add explict handling of BBHandledException for parsing failures

This removes some unnecessary tracebacks

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocooker: Ensure parsing failures stop the build
Richard Purdie [Wed, 22 Aug 2012 19:33:39 +0000 (20:33 +0100)]
cooker: Ensure parsing failures stop the build

Currently parsing failures still allow bitbake to continue on and try
and execute a build. This is clearly a bad idea and this patch adds in
more correct error handling and stops the build.

The use of sys.exit is nasty but this patches other usage in this function
so is at least consisent and its better than the current situation of
trying to execure a half parsed set of recipes. There are probably better
ways this could be improved to use to stop the build.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoutils.py: Once we've printed errors messages, suppress the backtrace
Richard Purdie [Wed, 22 Aug 2012 19:02:39 +0000 (20:02 +0100)]
utils.py: Once we've printed errors messages, suppress the backtrace

Adding a generic backtrace to better_compile and better_exec is pointless,
therefore reraise the exception as a bb.BBHandledException so the
generic code doesn't confuse the user even more.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoutils.py: Try harder to extract good traceback information by querying the datastore
Richard Purdie [Wed, 22 Aug 2012 19:01:12 +0000 (20:01 +0100)]
utils.py: Try harder to extract good traceback information by querying the datastore

Currently as soon as execution passes outside the code fragment being
executed by better_exec, we don't get any good traceback information,
just a likely obscure reference to some function name which may
or may not be identifiable.

This patch adds code to query the datastore if present, allowing a more
meaningful back trace to be displayed in many cases.

[YOCTO #2981]

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocooker: fix traceback when using -b with skipped recipe
Paul Eggleton [Wed, 22 Aug 2012 15:14:38 +0000 (16:14 +0100)]
cooker: fix traceback when using -b with skipped recipe

If a recipe is skipped during parsing for whatever reason, check and
report this as an error rather than trying to use the data that is sent
back and failing.

Fixes [YOCTO #2976].

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoevent/ast: Use better_exec instead of simple_exec
Richard Purdie [Tue, 21 Aug 2012 16:25:07 +0000 (16:25 +0000)]
event/ast: Use better_exec instead of simple_exec

This improves the stacktraces dumped by bitbake when for example anonymous
python functions fail.

Also default to passing code strings to better_exec to match the behaviour of
simple_exec to aid the transition.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoui: Improve error message if bitbake cannot import python curses module
Konrad Scherer [Tue, 21 Aug 2012 21:05:58 +0000 (16:05 -0500)]
ui: Improve error message if bitbake cannot import python curses module

On some SuSE systems, the curses python module is not installed by default.
Instead of a python failure, we want a nicer error message.

(On SuSE systems the package is typically python-curses.)

Signed-off-by: Konrad Scherer <Konrad.Scherer@windriver.com>
Signed-off-by: Jeff Polk <jeff.polk@windriver.com>
Reword commit message, rebase to latest bitbake.

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobitbake-layers: use parsable title for show-appends
Paul Eggleton [Tue, 21 Aug 2012 17:11:47 +0000 (18:11 +0100)]
bitbake-layers: use parsable title for show-appends

Print title text surrounded by === as we do with other subcommands so
that you can filter it out in scripts. (You will also want to filter out
lines starting with "WARNING:" if you just want the list).

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobitbake-layers: avoid full parse for show-layers subcommand
Paul Eggleton [Tue, 21 Aug 2012 17:11:46 +0000 (18:11 +0100)]
bitbake-layers: avoid full parse for show-layers subcommand

We don't actually need to parse all recipes just to show the configured
layers, so just parse the configuration instead.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobitbake/cooker: minor refactor of parseConfiguration()
Paul Eggleton [Tue, 21 Aug 2012 17:11:45 +0000 (18:11 +0100)]
bitbake/cooker: minor refactor of parseConfiguration()

All calls to parseConfiguration set self.status and call
self.handleCollections() afterwards, so just do this inside
parseConfiguration() itself.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agodata_smart: skip all interal (underscored) flags in getVarFlags
Ross Burton [Tue, 21 Aug 2012 16:16:26 +0000 (17:16 +0100)]
data_smart: skip all interal (underscored) flags in getVarFlags

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agodata-smart: rename the 'content' internal variable to '_content'
Ross Burton [Tue, 21 Aug 2012 16:16:25 +0000 (17:16 +0100)]
data-smart: rename the 'content' internal variable to '_content'

Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoHob wizard-like reorganization
Cristian Iorga [Tue, 21 Aug 2012 15:21:23 +0000 (18:21 +0300)]
Hob wizard-like reorganization

Hob is now more context-sensitive regarding
user changes/options. Also, the workflow have been
streamlined and resembles more of a wizard.
Beautified some hardcoded values.
Fixed typo.
New streamlined Image Configuration page.
Build and/or Edit image buttons presence
is context sensitive.
Recipes and packages tabs selected automatically
based on custom image or pre-defined target image
(included or all).
Context sensitive Back button.

Fixes [YOCTO 2165]

Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobitbake: hob: create console log directory before writing
Paul Eggleton [Tue, 21 Aug 2012 15:07:37 +0000 (16:07 +0100)]
bitbake: hob: create console log directory before writing

The default value of BB_CONSOLELOG in OE-Core now points to a directory
that might not exist prior to the first build, so ensure it is created
first.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobuild.py: Correct package to recipe in TaskBase events to use consistent terminology
Richard Purdie [Fri, 17 Aug 2012 13:53:16 +0000 (14:53 +0100)]
build.py: Correct package to recipe in TaskBase events to use consistent terminology

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2/git: Work around git confusion between foo.git and foo repositories
Richard Purdie [Sat, 18 Aug 2012 15:15:18 +0000 (15:15 +0000)]
fetch2/git: Work around git confusion between foo.git and foo repositories

If you have foo and foo.git in GITDIR, the two can end up being confused
by git with some horrible union of the two being cloned. This adds
a workaround to avoid this happening until git 1.7.9.2 onwards is
common enough for this to be removed. We use a symlink to hide
the directories we don't want git to know about.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoreally remove knotty2
Richard Purdie [Fri, 17 Aug 2012 13:52:40 +0000 (14:52 +0100)]
really remove knotty2

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobin/bitbake: Remove knotty2 from the list of valid interfaces
Richard Purdie [Thu, 16 Aug 2012 14:15:49 +0000 (15:15 +0100)]
bin/bitbake: Remove knotty2 from the list of valid interfaces

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobitbake: build.py: Add stampdir argument to cached_mtime_noerror
Andrei Gherzan [Thu, 16 Aug 2012 11:21:33 +0000 (14:21 +0300)]
bitbake: build.py: Add stampdir argument to cached_mtime_noerror

After commit 2718537b4b04eb3d80ab4d74171b58e7b8dd68b8 (bitbake:
build.py: Only execute mkdirhier if stampdir doesn't exist) build
failes as cached_mtime_noerror needs an argument - stamp dir. This
argument was forgotten.

Signed-off-by: Andrei Gherzan <andrei.gherzan@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoknotty: Fold knotty2 into knotty and make it the default
Richard Purdie [Wed, 15 Aug 2012 16:50:22 +0000 (17:50 +0100)]
knotty: Fold knotty2 into knotty and make it the default

There is no good reason knotty2 shouldn't be the default now. If you need
the old behaviour, just pipe the output through cat as non-interactive
terminals get the old output.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoknotty2: Handle long lines of text and terminal window size changes
Richard Purdie [Wed, 15 Aug 2012 16:00:40 +0000 (17:00 +0100)]
knotty2: Handle long lines of text and terminal window size changes

Long lines of text which wrapped on the terminal corrupted the output shown by
knotty2. This patch catches such errors by becomming aware of the terminal size.

It also catches terminal window size change events and adapting to those
changes using a signal handler.

Based on a patch from Jason Wessel with several tweaks and enhancements
such as use of chained signal handlers and covering all output messages.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobuild.py: Only execute mkdirhier if stampdir doesn't exist
Richard Purdie [Wed, 15 Aug 2012 15:57:47 +0000 (16:57 +0100)]
build.py: Only execute mkdirhier if stampdir doesn't exist

I noticed this was showing up on profile logs as a sigificant time user
in "bitbake bash" when bash was already built. It reduces the time from
5.2 to 4.5 seconds in my test environment.

We make use of the parser's mtime cache as once a directory exists,
we can assume it continues to exist and this avoids syscalls.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobb/ui/uihelper.py: Ensure task current and total numbers are updated for setscene...
Richard Purdie [Wed, 15 Aug 2012 15:59:57 +0000 (16:59 +0100)]
bb/ui/uihelper.py: Ensure task current and total numbers are updated for setscene events too

This avoids task (0 of 0) type messages being displayed during setscene by knotty2.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoknotty: Chain the WINCH signal handler in progressbar so any other handler is preserv...
Richard Purdie [Wed, 15 Aug 2012 15:58:46 +0000 (16:58 +0100)]
knotty: Chain the WINCH signal handler in progressbar so any other handler is preserved and called

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agodata_smart: Fix unanchored regexp causing strange parsing issue
Richard Purdie [Fri, 3 Aug 2012 09:24:32 +0000 (10:24 +0100)]
data_smart: Fix unanchored regexp causing strange parsing issue

If this regular expression is unanchored, it would accept strings like:

do_install_append1
do_install_appendsomelongstring

and treat them like they were do_install_append. Clearly this isn't desirable.
Only one instance of this type of issue was found in OE-Core and has been fixed
so correcting the regexp should be safe to do.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoFix script location after mishandled merge
Richard Purdie [Fri, 3 Aug 2012 08:04:07 +0000 (09:04 +0100)]
Fix script location after mishandled merge

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocooker: fix OVERRIDES in BB_SIGNATURE_HANDLER (to override value from DISTRO.conf...
Martin Jansa [Thu, 2 Aug 2012 16:52:15 +0000 (18:52 +0200)]
cooker: fix OVERRIDES in BB_SIGNATURE_HANDLER (to override value from DISTRO.conf from local.conf)

We really need to pass the finalised data store into the parser init function
(and hence the siggen init function). This ensures any value changes get
passed into the correct code.

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2/local: Add search paths to the debug log to improve log usefullness
Richard Purdie [Thu, 2 Aug 2012 20:48:31 +0000 (20:48 +0000)]
fetch2/local: Add search paths to the debug log to improve log usefullness

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2/local.py: Provide better debug output when fetch of a local file fails
Richard Purdie [Thu, 2 Aug 2012 20:41:02 +0000 (20:41 +0000)]
fetch2/local.py: Provide better debug output when fetch of a local file fails

When a fetch failure occurs for a local file, this patch ensures we print the
locations searched making it easier for the user to debug the problem.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agorunqueue.py: Clean up runqueue exception catching in the normal task failure case
Richard Purdie [Thu, 2 Aug 2012 20:40:36 +0000 (20:40 +0000)]
runqueue.py: Clean up runqueue exception catching in the normal task failure case

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agocontrib/dump_cache.py: dump recipe -> packages mapping from bb_cache.dat
Robert Yang [Wed, 1 Aug 2012 07:18:18 +0000 (15:18 +0800)]
contrib/dump_cache.py: dump recipe -> packages mapping from bb_cache.dat

Add the dump_cache.py to dump the "recipe -> packages" mapping for
target recipes form bb_cache.dat:

* Usage:
  dump_cache.py bb_cache.dat

* The format is:
    recipe_path pn pv packages
  For example:
    /path/to/gzip_1.5.bb gzip 1.5 gzip-dbg gzip-staticdev gzip-dev gzip-doc gzip-locale gzip

* Only save the mapping for the target recipe

* We can extend this to dump other informations when needed.

* Put this script to bitbake/contrib/ (not to oe-core) is because it
  needs the bb.cache.

[YOCTO #2741]

Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agobitbake-layers: fix help for show-overlayed to match behaviour
Paul Eggleton [Tue, 24 Jul 2012 11:41:24 +0000 (12:41 +0100)]
bitbake-layers: fix help for show-overlayed to match behaviour

show-overlayed actually lists the same recipe in multiple layers
regardless of priority, so change the help description to match.

Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoHob: change the Build failure scenario as ui design
Liming An [Thu, 2 Aug 2012 11:06:02 +0000 (19:06 +0800)]
Hob: change the Build failure scenario as ui design

As ui design 'build-fail-spec.pdf' to change the GUI

[YOCTO #2183]

Signed-off-by: Liming An <limingx.l.an@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoHob: log for Hob and allow users to show logs after successful build
Liming An [Thu, 2 Aug 2012 11:06:01 +0000 (19:06 +0800)]
Hob: log for Hob and allow users to show logs after successful build

If users build images in Hob, record logs and allow users to retrieve the lo
after successful build.

The logs are generated if and only if:
      - users do "just bake"
      - users do "build image" after "build packages"
      - users do "build packages" only

[YOCTO #1991]

Signed-off-by: Shane Wang <shane.wang@intel.com>
Signed-off-by: Liming An <limingx.l.an@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2: print checksums when they are different then expected
Martin Jansa [Tue, 31 Jul 2012 23:33:41 +0000 (01:33 +0200)]
fetch2: print checksums when they are different then expected

* in form which can be copied into a recipe without modification
* like oe-classic did since:
  http://git.openembedded.org/openembedded/commit/?id=68abc465559a68e9201c9525be3c4acc6297eaed
* it shows them in right form when they are missing completely, but in
  more verbose form when different
* it needs to print that only when checksums were requested, e.g.
  fetching from sstate mirror sets both md5mismatch and sha256mismatch,
  but your checksums shouldn't be shown

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2: handle broken symlinks in local mirror handling
Christopher Larson [Tue, 31 Jul 2012 22:20:14 +0000 (15:20 -0700)]
fetch2: handle broken symlinks in local mirror handling

If a file:// mirror is being used, the fetcher will create a symlink to the
local file. However, if the local file gets removed, that link will be dead,
and os.path.exists() returns False in that case, so it tries and fails to
recreate the link. Now we unlink such a dead link if it exists.

Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agoknotty: ensure that directory for BB_CONSOLELOG exists before using it
Martin Jansa [Mon, 30 Jul 2012 11:11:46 +0000 (13:11 +0200)]
knotty: ensure that directory for BB_CONSOLELOG exists before using it

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agofetch2/__init__.py: Add NoChecksumError exception
Mark Hatle [Sat, 28 Jul 2012 01:26:38 +0000 (20:26 -0500)]
fetch2/__init__.py: Add NoChecksumError exception

Without the new exception, when the system is configured to use premirrors,
but not allow network access (via BB_NO_NETWORK), when a recipe was lacking a
checksum the wrong error message(s) were being generated.

Instead of complaining about trying to perform network access, if the system
was able to find the item in the premirror, it should inform the user of the
lack of checksums, and the two SRC_URI fields they should use to update their
recipe.

Signed-off-by: Mark Hatle <mark.hatle@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agohob: deal event DiskFull
Kang Kai [Fri, 27 Jul 2012 08:38:03 +0000 (16:38 +0800)]
hob: deal event DiskFull

Part of [Yocto #2168]

When bitbake runqueue is teminated by disk monitor, it will send event
DiskFull. Update to handle it.

Signed-off-by: Kang Kai <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
8 years agomonitordisk: fire event DISKFULL when terminate build
Kang Kai [Fri, 27 Jul 2012 08:38:02 +0000 (16:38 +0800)]
monitordisk: fire event DISKFULL when terminate build

Part of [Yocto #2168]

Add a event DiskFull to descript the termination by disk monitor.

Update check() to fire the event DiskFull when terminates the build.
This could help UIs to deal this scenario and show more information to
end user.

Signed-off-by: Kang Kai <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>