The "Speeding Up the Build" section presents a list of items the
user can consider when they want to tweak the build speed. One of
the items is the three variables BB_NUMBER_THREADS, BB_NUMBER_PARSE_THREADS,
and PARALLEL_MAKE. It was noted that the system figures out
the optimal settings for these now and it is pretty useless to
override the settings. However, the introducing sentence to the
list is still valid due to other considerations that go beyond these
three variables. So, to update the text, I have re-emphasized the
role of these three variables for quickening the build time.
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
(From yocto-docs rev: 84610550cef66d0aa7478bcf95ad97b081709b2f)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Updated some text to indicate that you should put patch files
in $BP, $BPN, or "files." Also, added more on how the patch
files suffix is (i.e. accounting for compressed suffix names
like diff.bz).
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
(From yocto-docs rev: d9e23da24e9d86ec5e2b9d4067667478f3e56ba1)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
In the "Fetching Code" section, some loose description of how the
build system reacts to the two checksums in a recipe existed.
Robert asked some questions about how the behavior would be if
just a single, correct checksum existed and if the two existed but
only one proved correct. Turns out, if you supply a single one then
the build will not throw an error or warning. If you supply two
but one is incorrect, then the build will throw an error or
warning. I updated the text to be technically correct without
going into huge details.
(From yocto-docs rev: f49a744821d9326314803d5003158bd402e74e0d)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some minor edits to the original patch. These edits were caught by
Robert P. J. Day.
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
(From yocto-docs rev: b50a2e44eccd50166351611f17c643cbbc60f8b8)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Placed a reference for more information on the variable into the
section of the dev-manual that explains how to choose a device
manager.
(From yocto-docs rev: 3f9b1329fe4794af03d673323d11558e85d9b097)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This patch was modified to fit in with the style of the book.
(From yocto-docs rev: 7c6c7ac573f92bfc844977e0d7a97ca4aac9bc70)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Bob Cochran pointed out that the skeleton recipe shown in the
example is not really what is created by newbb.vim. I updated the
example to match.
(From yocto-docs rev: 2dfa712522aec5204b9e739b6060109090280011)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This section was a bit confusing on the use of "_prepend" as associated
with the FILESEXTRAPATHS statement. Especially in the final note
where it indicated you don't need "_prepend" if your append file
is not adding any new files. I updated the note to be more specific.
Another small change I made was to replace the Crown Bay BSP referenced
in the section with the Emenlow BSP, which has a longer life span than
CB.
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
(From yocto-docs rev: 35bdc8031bc4ecf1ea321b0cafade97eeaa56513)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When you "make" the mega-manual, there were four tags across the
various manuals where the same string "id='intro'" occurred. This
causes warnings during the make process. I updated the strings to
be unique. There are no cross-references to these so no collateral
issues were caused.
(From yocto-docs rev: d0e770e0ad797031a75e2373a41723408b4b051c)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
I updated the following variables based on slight review changes
from Paul Eggleton:
EXTRA_AUTORECONF - changed autoconf class to autotools
KERNEL_IMAGE_MAXSIZE - Wording tweak
AUTO_LIBNAME_PKGS - Wording tweak
DEBIANNAME - [doc] string updated
TCLIBCAPPEND - Second clarifying sentence in first para added
PACKAGEGROUP_DISABLE_COMPLEMENTARY - [doc] string updated
(From yocto-docs rev: ff5c765ffab1cdf6fc8a007cebdf7114432f65ef)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The section "Customizing Images Using Custom IMAGE_FEATURES
and EXTRA_IMAGE_FEATURES" needed a bit more clarification on
the image features. The text pretty much passed off image
features as all mapping into package groups, which is not true.
Some, like "debug-tweaks for example, are more like general
configurations. I updated the section accordingly.
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
(From yocto-docs rev: 7dd55ff0bd6f7923e22695ed7323df5cfe17d1d4)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This note provides a bit of a warning or tip regarding naming
BSP layers. The letter of the law dictates that you don't
need to prepend a BSP layer name with the string "meta-".
However, scripts exist in OE where the naming convention is
assumed.
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
(From yocto-docs rev: 0375349fdd3cd5ff8a54db8124d5cdc02133ff12)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixes [YOCTO #7029]
I was missing the "sysroots" part of the path. I added this
to the pathname.
(From yocto-docs rev: 077efc8063ed01ef4067f7e71912dd52c1f6a4a6)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Crown Bay is short on life so I swapped it out for the eMenlow.
(From yocto-docs rev: 6d5394c12a374463cb1d5215af88178b63a5b8c0)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The yocto-bsp and yocto-kernel tool output for the examples
were updated to be current.
(From yocto-docs rev: a05b1df35c7518027b1eb73f9ea2b6a3d00dca43)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Added descriptions for the following variables:
EXTRA_AUTORECONF
KERNEL_IMAGE_MAXSIZE
UPDATERCPN
AUTO_LIBNAME_PKGS
DEBIANNAME
DEBIAN_NOAUTONAME
TCLIBCAPPEND
PACKAGEGROUP_DISABLE_COMPLEMENTARY
CLEANBROKEN
DISTRO_CODENAME
BASE_WORKDIR
DISTRO_FEATURES_DEFAULT
DISTRO_FEATURES_LIBC
Part of adding new variables is checking to see if a previously
undocumented variable is used elsewhere in the manual. One
occurrence was spotted and turned into a link for this list.
(From yocto-docs rev: bb75387b7aa4fcb0b594d300772952d81a651644)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
By default, making a file will not include the PDF version of the
manual. Some manuals you can't do a PDF version anyway (i.e.
yocto-project-qs and mega-manual). The PDF versions do not need
to be part of the mainstream make process. If someone wants these
they can take the extra step of doing 'make pdf DOC=<manual>'.
(From yocto-docs rev: 16f605350adb1b252b94e921faf282d0634cefb6)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
I need to get back to how it was working and then fix the table
box issue from there. Using the 1.76.1 stylesheets makes the
build too long for the manuals and sometimes the XSL sheets do
not load.
(From yocto-docs rev: 7bc502fa069fdc4aebf5705513bb82127716859f)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This section was pointing to a poor example in meta. I substituted
in a better example and rewrote the section to flow better around
that example and the final fabricated one.
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
(From yocto-docs rev: c7f07a2993d5c1b8e0c8b7da0e9d4c28242a5e26)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There was a sample string in this tree that ended in "ro" when it
should have been "r0" (as in release 0). Anytime you update a figure,
the modified file must go in both the "figures" directory of both
the manual (dev-manual in this case) and in the mega-manual.
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
(From yocto-docs rev: 8e75cc23fbf03f355fade416a12b7c8fe1e656a5)
Signed-off-by: Scott Rifenbark <scott.m.rifenbark@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Toaster needs to properly prompt the user and display
actionable help texts when the database is void of
meaningful information. This patch brings in the
"empty" states for pages.
[YOCTO #6755]
(Bitbake rev: 7cfe279bf77b59d5cbd20d8a93e1d33279bebc20)
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This brings it into line with the all layers and targets pages
(Bitbake rev: 7c862ae4e3c3a00f2b37c2272007ce64304359bd)
Signed-off-by: Michael Wood <michael.g.wood@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
From the bug: The layer source reveals too much of the implementation model,
without explaining what is going on behind the scenes. It will confuse more
than inform, so let's get rid of it.
[YOCTO #7302]
(Bitbake rev: 7fee378849c9518a4560aa5be65fca9644c03048)
Signed-off-by: Michael Wood <michael.g.wood@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Target input field needed to support both selecting suggestions and
typing targets, since Toaster doesn't always know about all the targets
provided by the layers in the project.
[YOCTO #7187]
(Bitbake rev: d753db7a18eb9218016e586171d2a08298a0b05d)
Signed-off-by: Michael Wood <michael.g.wood@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Use the new layer_version api so that we return the vcs reference whether
that be the commit hash, branch name or the upstream branch to which the
layer is associated with.
[YOCTO #7192]
(Bitbake rev: fa6e6d5b18c8c6231ae8ac499d332298f551fd7d)
Signed-off-by: Michael Wood <michael.g.wood@intel.com>
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This patch fixes the git checkout command that would fail to
properly rebase the working tree in some cases.
[YOCTO #7309]
(Bitbake rev: 25640851c8f8459741d2c423b2bb54f9f59a536a)
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This patch brings in reading the server log if the server
fails to start. This is the first step in displaying the errors
in a manner that is actionable by the user.
[YOCTO #7285]
(Bitbake rev: 7f47b67bccff55673a827dbcccc767af4459c933)
Signed-off-by: Alexandru DAMIAN <alexandru.damian@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Avoid using keyup event to trigger form validation as this won't fire
for events such as pasting or autofill from the browser.
[YOCTO #7292]
(Bitbake rev: 2060be8d31baa0cd302aa8dc0b6d179df853ee99)
Signed-off-by: Michael Wood <michael.g.wood@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Sometimes you do not want certain packages to be installed when
installing complementary packages, e.g. when using dev-pkgs in
IMAGE_FEATURES you may not want to install all packages from a
particular multilib. This introduces a new PACKAGE_EXCLUDE_COMPLEMENTARY
variable to allow specifying regexes to match packages to exclude.
(From OE-Core rev: d4fe8f639d87d5ff35e50d07d41d0c1e9f12c4e3)
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
* we were stripping too much when stripping recipe name from line like this:
ERROR: Task 12016 (/some/patch/something.dot.bar.bb, do_fetch) failed with exit code '1'
where the recipe name contains dots and doesn't end with _<version>.bb
(From OE-Core rev: 8c9a25ae70d249b823ab2b0385d539eb8bbc1374)
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a check to the SDK so that it only runs on systems with kernel versions
it supports.
[YOCTO #6856]
(From OE-Core rev: b4caa8085aa15674162ff5135b13409998db9510)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
bitbake will currently 'selecting bluez4 to satisfy runtime
libasound-module-bluez due to PREFERRED_PROVIDER_bluez4 = bluez4'
which in the case of bluez5 isn't correct.
This slightly unusual construct avoids this.
Ultimately this is a bitbake issue that needs fixing in
a better way but this means we can merge the bluez5 changes
until bitbake gets fixed.
(From OE-Core rev: ef41f4b91d65f87850edd6cc56ca37d2ecb56378)
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
connman currently supports both bluez4 and bluez5 without distinguishing
them at configuration time.
(From OE-Core rev: 420ab90167e434d6972702c44271104881f08fc9)
Signed-off-by: Peter A. Bigot <pab@pabigot.com>
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
