Skip to content

L
linux
Switch branch/tag
  1. 13 Jan, 2010 1 commit
    • Paul Mundt's avatar
      sh: Move over to dynamically allocated FPU context. · 0ea820cf
      Paul Mundt authored 
      This follows the x86 xstate changes and implements a task_xstate slab
cache that is dynamically sized to match one of hard FP/soft FP/FPU-less.

This also tidies up and consolidates some of the SH-2A/SH-4 FPU
fragmentation. Now fpu state restorers are commonly defined, with the
init_fpu()/fpu_init() mess reworked to follow the x86 convention.
The fpu_init() register initialization has been replaced by xstate setup
followed by writing out to hardware via the standard restore path.

As init_fpu() now performs a slab allocation a secondary lighterweight
restorer is also introduced for the context switch.

In the future the DSP state will be rolled in here, too.

More work remains for math emulation and the SH-5 FPU, which presently
uses its own special (UP-only) interfaces.
Signed-off-by: default avatarPaul Mundt <lethal@linux-sh.org>
      0ea820cf
  3. 12 Jan, 2010 11 commits
  5. 08 Jan, 2010 1 commit
    • Paul Mundt's avatar
      sh: consolidate atomic_cmpxchg()/atomic_add_unless() definitions. · 8c0b8139
      Paul Mundt authored 
      The LL/SC and IRQ versions were using generic stubs while the GRB version
was just reimplementing what it already had for the standard cmpxchg()
code. As we have optimized cmpxchg() implementations that are decoupled
from the atomic code, simply falling back on the generic wrapper does the
right thing. With this in place the GRB case is unaffected while the
LL/SC case gets to use its optimized cmpxchg().
Signed-off-by: default avatarPaul Mundt <lethal@linux-sh.org>
      8c0b8139
  7. 06 Jan, 2010 1 commit
  9. 05 Jan, 2010 2 commits
  11. 04 Jan, 2010 5 commits
  13. 02 Jan, 2010 3 commits
    • Matt Fleming's avatar
      sh: Move page table allocation out of line · 2a5eacca
      Matt Fleming authored 
      We also switched away from quicklists and instead moved to slab
caches. After benchmarking both implementations the difference is
negligible. The slab caches suit us better though because the size of a
pgd table is just 4 entries when we're using a 3-level page table layout
and quicklists always deal with pages.
Signed-off-by: default avatarMatt Fleming <matt@console-pimps.org>
      2a5eacca
    • Matt Fleming's avatar
      sh: Optimise flush_dcache_page() on SH4 · b4c89276
      Matt Fleming authored 
      If the page is not mapped into any process's address space then aliases
cannot exist in the cache. So reduce the amount of flushing we perform.
Signed-off-by: default avatarMatt Fleming <matt@console-pimps.org>
      b4c89276
    • Matt Fleming's avatar
      sh: Correct the PTRS_PER_PMD and PMD_SHIFT values · 3f5ab768
      Matt Fleming authored 
      The previous expressions were wrong which made free_pmd_range() explode
when using anything other than 4KB pages (which is why 8KB and 64KB
pages were disabled with the 3-level page table layout).

The problem was that pmd_offset() was returning an index of non-zero
when it should have been returning 0. This non-zero offset was used to
calculate the address of the pmd table to free in free_pmd_range(),
which ended up trying to free an object that was not aligned on a page
boundary.

Now 3-level page tables should work with 4KB, 8KB and 64KB pages.
Signed-off-by: default avatarMatt Fleming <matt@console-pimps.org>
      3f5ab768
  15. 31 Dec, 2009 1 commit
  17. 29 Dec, 2009 1 commit
    • Paul Mundt's avatar
      sh: Only provide a PCLK definition for legacy CPG CPUs. · 8152a74b
      Paul Mundt authored 
      As CPUs are migrated over to more fully-featured clock frameworks of
their own and off of the legacy CPG code, they no longer have any real
need for defining the PCLK value. The PCLK define in itself is already
fairly misleading, as many boards get their input clocks from different
sources, making this value fairly arbitrary anyways.

Outside of the legacy CPG clock framework, the only place where this
value is used is for deriving CLOCK_TICK_RATE, which we set back to the
legacy PIT value that it was before the PCLK definitions were added in
the first place.
Signed-off-by: default avatarPaul Mundt <lethal@linux-sh.org>
      8152a74b
  19. 24 Dec, 2009 14 commits
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7