Our yaffs filesystems reserve 5 blocks, use 1 for a checkpoint, and
may use 1 or 2 more for caching. Take these all into account when
computing the max image size.
The WebKit library link line is exceeding 128KB on sim-eng builds.
The path to sim-eng object files is slightly longer than device builds
because the object files live under the "host" directory.
This change truncates the "product" directory name to "pr", reducing the
command line by a few KB. This only affects sim-eng builds.
The real fix will need be to webkit (see internal bug 1917987), which
will eventually start failing on device builds if it continues to grow.
* changes:
Add the ability to turn on proguard for java targets, and to call aapt to generate a rules file for proguard for stuff from the manifest and other resources that need to be kept.
This is done by appending "mdpi" (medium density or 160 dpi, which is what
we've been using so far on dream etc.) to PRODUCT_LOCALES if no other dpi is
found there. (PRODUCT_LOCALES is then sent as a configuration parameter to
aapt to determine which resources should be included in the build.)
With this change, each build will now include only one density of any given
asset. For example, if there exists
drawable-hdpi/foo.png
drawable-mdpi/foo.png
...then the build will include *only* drawable-mdpi/foo.png, unless the
density is overridden in the build's own device spec:
PRODUCT_LOCALES += hdpi
...in which case the build will include only drawable-hdpi/foo.png.
Merge commit '1b6d0a65fc62901f2c18d00c556bd5da50f35584' into eclair
* commit '1b6d0a65fc62901f2c18d00c556bd5da50f35584':
allow prebuilts to specify LOCAL_CERTIFICATE so they can be resigned
If a prebuilt specifies a LOCAL_CERTIFICATE, stick it in the
PACKAGES.* collection so that the predexopt process can work on the
file (which only works if the new .apk can be signed).
I removed these lines because for some reason building under Windows
fails for me when they are present. However the build server breaks
so I'll rolling them back and I'll try to understand that later.
Merge commit '6929fb57a5afb90606feb8c35c8c2cda170adfe4'
* commit '6929fb57a5afb90606feb8c35c8c2cda170adfe4':
Fix issue #2048267: Run zipalign on all prebuilts
For host side test packages, the name of the jar file containing the tests is
declared in the packages Android.mk. The name of the make file needs to be
passed to the description generator so that it can pick up this information.
The makefile was not passing on the parameter containing the make file name.
Make BoardConfig.mk store the size of the partition rather than the
maximum size of the image that can be flashed there, because the
function used to do the conversion isn't available when BoardConfig.mk
is read any more.
Change things so that when $(LOCAL_PREBUILT_STRIP_COMMENTS) is
nonempty, we copy the source file using sed to strip out all the "#
line"-style comments and blank lines, saving considerable system image
space in the case of some wifi configurations.
Bug: 2036961
This cuts the make startup time by about 30 seconds. Python is faster
than bash in this case, and also we can now supply multiple directories
to prune, and skip the .repo directory, which is, uh, big.
This is from my mac laptop:
$ time build/tools/findleaves.sh --prune="./out" . Android.mk > /dev/null
real 0m29.186s
user 0m0.550s
sys 0m5.897s
$ time build/tools/findleaves.py --prune="./out" . Android.mk > /dev/null
real 0m4.701s
user 0m0.645s
sys 0m1.294s
$ time build/tools/findleaves.py --prune="./out" --prune="./.repo" . Android.mk > /dev/null
real 0m0.176s
user 0m0.094s
sys 0m0.080s
Doug Zongker