Skip to content


Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Commits on Jun 26, 2013
  1. @ahmedammar
Commits on Jun 25, 2013
  1. @ahmedammar
Commits on Jun 18, 2013
  1. @ahmedammar

    ESS-165: increase DMA zone

    ahmedammar authored
  2. @ahmedammar
Commits on Jun 10, 2013
  1. @ahmedammar
Commits on Jun 6, 2013
  1. amd-gpu: correct sharedmem allocation

    Matt Sealey authored
    Merge/tree mishap caused this code to go missing.
    Signed-off-by: Matt Sealey <>
  2. amd-gpu: remove aperture concept from the memory manager

    Matt Sealey authored
    Now everything is one big memarena instead of a separate PHYS and EMEM,
    and instead of having to micromanage the split between Z160 (always PHYS
    here) and GPU (99% of it's allocations come out of EMEM), we give blocks
    from the same memory pool.
    This should effect some more efficient allocation strategies and allow
    Xorg acceleration to proceed with much less movement of data (when it
    evicts for low memory in PHYS, it moves it to userspace and uses pixman
    to render everything, which is CPU intensive) at the potential cost of
    leaving slightly less memory around for the 3D GPU, but userspace apps
    using the GPU should be doing better with texture management so as not
    to soak up >32MiB of reserved GPU memory needlessly.
    Moving away from gsl_memmgr to genalloc is not desirable right now since
    the driver requires multiple allocation alignment orders and genalloc can
    only really guarantee a minimum alignment (i.e. size of allocation
    represented by each bit in the allocation bitmap).
    Moving to a dma_alloc_coherent/writecombine-only model also poses a
    problem since significant amounts of code expects to be able to find the
    "base" GPU memory address and with individual allocations instead of a
    pooled area, there is no such thing.
    Signed-off-by: Matt Sealey <>
Commits on May 22, 2013
  1. ipu: don't follow interface pixfmt again since 32-bit is problematic

    Matt Sealey authored
    Signed-off-by: Matt Sealey <>
Commits on May 21, 2013
  1. @ahmedammar

    amd-gpu: rework apertures

    ahmedammar authored
    * increase from 64MB to 96MB total memory
    * after some testing it seems that PHYS can be quite small compared
      to EMEM, therefore make the aperture split 16MB/80MB
Commits on May 5, 2013
  1. @ahmedammar

    amd-gpu: increase MEM1 area some more for latest eglsink

    ahmedammar authored
    - fixes crashes with full-screen 720p eglsink
Commits on Mar 1, 2013
  1. Revert "ARM 5784/1 fix early boot machine ID mismatch error display"

    Matt Sealey authored
    Mistakenly applied this to 2.6.31 all those years ago while doing
    experiments on later kernels. Actually it depends on several other
    patches which never got applied.
    This reverts commit 9dda27b.
    Signed-off-by: Matt Sealey <>
Commits on Jan 20, 2013
  1. @ahmedammar

    amd-gpu: increase MEM1 area

    ahmedammar authored
    - fixes issue where allocations fail after compiz started.
  2. @ahmedammar

    Merge branch 'amd-gpu' of git://

    ahmedammar authored
Commits on Dec 11, 2012
  1. ENGR00162464 update pm4 microcode: pm4_microcode_r18_20111020.a.inl.rel

    Richard Zhao authored Matt Sealey committed
    It fix gpu hang.
    Signed-off-by: Richard Zhao <>
  2. config: use preemptible rcu

    Matt Sealey authored
  3. config: disable BZIP2 ramdisk compression

    Matt Sealey authored
    Nobody uses it, it's slow to compress, slow to decompress, we support
    GZIP and LZMA anyway.
  4. config: disable sisusbvga (depends on vga_8x16) and enable the fixed_…

    Matt Sealey authored
    …8x16 font
  5. font: add the Fixed 8x16 console font to the kernel.

    Matt Sealey authored
    It's a bit like Terminus, but more rounded and less harsh on the eyes. Ubuntu has shipped
    it as their default console font for years, before they switched to the "Ubuntu" font. Putting
    this in the kernel has practically no benefit except to reduce the amount of "glitching"
    swapping the built-in kernel console font for a new one halfway through boot when you see the
    kernel output on a screen.
    Note to gain the benefit of it not swapping the font around again when "consolefont" or
    similar service starts, you need to set your font to Uni3-Fixed16.psf.gz or something very
  6. amd-gpu: use kzalloc instead of kmalloc/memset

    Matt Sealey authored
  7. ENGR00160722 amd gpu: add best fit algorithm to memory manager

    Xinyu Chen authored Matt Sealey committed
    The AMD gpu driver memory manager is stupid to allocating
    a memory block from the first free block. Neither check the
    best fit free block. Here just add a little codes to
    let allocation to find a best size fitted free block.
    To avoid memory fragement.
    Signed-off-by: Xinyu Chen <>
  8. ENGR00162193 amd gpu: fixed random gsl_kmod failure

    Richard Liu authored Matt Sealey committed
    increase max process number to fix gsl_kmod open issue
    Signed-off-by: xianzhong <>
    Signed-off-by: Richard Liu <>
  9. config: re-enable PREEMPT

    Matt Sealey authored
  10. BFS: back out BFS scheduler for stability reasons

    Matt Sealey authored
    It works faster sometimes, other times not.. the code is so old now
  11. amd-gpu: more sane way of doing an RB write

    Matt Sealey authored
  12. amd-gpu: ringbuffer polling/timestamp buffer alignment fix

    Matt Sealey authored
    According to Qualcomm it can be at a 4-byte boundary, but for using MMU
    it needs to be at a page boundary
  13. amd-gpu: don't use writecombining for remapping the sharedmem space e…

    Matt Sealey authored
    …ven with MMU
  14. amd-gpu: barriers in ringbuffer submit to ensure that the buffer is p…

    Matt Sealey authored
    …osted first
  15. amd-gpu: memory barrier for safety after issueibcmds sync mem

    Matt Sealey authored
    Makes sense (and is from Qualcomm's driver) since we cannot guarantee
    the memory writes have finished posting before we submit them to the
Something went wrong with that request. Please try again.