| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
| |
The patch gives more detailed description of cortex-A9's
pipeline. It in general improves performance on Cortex-A9.
Use -mtune=cortex-a9 to take advantage of this feature.
Change-Id: I803daa6631272a0dae3514b378b9f611b8fd2b54
|
|
|
|
|
|
|
|
|
| |
The problem was caused by backporting a patch from upstream without
its necessary dependencies. This patch changes the implementation of
cbranchqi4 so that we keep the optimization without generating
bad code.
Change-Id: Icbaf443ac6e85774cb664036941bf829c5e4c31e
|
|
|
|
|
|
|
|
|
| |
Fix a bad interaction of inliner x return slot optimization x
named value optimization
Replace tst instruction with lsls for a single bit test
Change-Id: I30c212d9e99c79eae182da66b97d9d9f4d738faf
|
|
|
|
|
|
|
|
|
| |
From now on, the general Android toolchain will be arm-linux-androideabi.
arm-eabi toolchain built from this source does not contain Android specific
configurations and will only be used to build Android kernel.
Read recent changes in toolchain/build/README for more details.
Change-Id: I7526576949b55809da5c24d34e1df884301c05b2
|
|
The source is based on fsf gcc-4.4.3 and contains local patches which
are recorded in gcc-4.4.3/README.google.
Change-Id: Id8c6d6927df274ae9749196a1cc24dbd9abc9887
|