<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Thu, Mar 31, 2016 at 2:59 AM, Samuel Iglesias Gonsálvez <span dir="ltr"><<a href="mailto:siglesias@igalia.com" target="_blank">siglesias@igalia.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hello,<br>
<br>
This is the second version of this patch series [0].<br>
<br>
In case you prefer a repository, it is available here [1]:<br>
<br>
$ git clone -b nir-bit-size-fixes-2.0 <a href="https://github.com/Igalia/mesa.git" rel="noreferrer" target="_blank">https://github.com/Igalia/mesa.git</a><br>
<br>
Thanks,<br>
<br>
Sam<br>
<br>
[0] <a href="https://lists.freedesktop.org/archives/mesa-dev/2016-March/110465.html" rel="noreferrer" target="_blank">https://lists.freedesktop.org/archives/mesa-dev/2016-March/110465.html</a><br>
[1] <a href="https://github.com/Igalia/mesa/tree/nir-bit-size-fixes-2.0" rel="noreferrer" target="_blank">https://github.com/Igalia/mesa/tree/nir-bit-size-fixes-2.0</a><br>
<br>
Connor Abbott (13):<br>
<span class="">  nir/from_ssa: adapt to different bit sizes<br>
  nir/locals_to_regs: adapt to different bit sizes<br>
</span><span class="">  nir/lower_vec: adapt to different bit sizes<br>
  nir/glsl_to_nir: support doubles<br>
</span>  nir/print: add support for printing doubles and bitsize<br>
<span class="">  nir: handle doubles in nir_deref_get_const_initializer_load()<br>
  nir/instr_set: handle 64-bit bit-sizes<br>
  nir/split_var_copies: handle doubles<br>
  nir: add support for d2f and f2d<br>
  nir: add support for (un)pack_double_2x32<br>
  nir: don't try to scalarize unpack_double_2x32<br>
  nir: add split versions of (un)pack_double_2x32<br>
  nir: add a pass for lowering (un)pack_double_2x32<br>
<br>
</span>Iago Toral Quiroga (7):<br>
<span class="">  nir/lower_to_source_mods: Handle different bit sizes<br>
</span><span class="">  nir/lower_load_const_to_scalar: suppport doubles and multiple bit<br>
    sizes<br>
  nir/glsl_to_nir: set the bit-size in the result of evaluate_rvalue<br>
</span><span class="">  nir/glsl_to_nir: set bit_size on ssbo_load result<br>
  nir: add d2i, d2u, d2b opcodes<br>
  nir: add i2d and u2d opcodes<br>
  nir/lower_to_source_mod: Skip unsafe operations<br>
<br>
</span>Samuel Iglesias Gonsálvez (3):<br>
  nir: add bit_size info to nir_ssa_undef_instr_create()<br>
  nir: add bit_size info to nir_load_const_instr_create()<br>
<span class="">  nir: verify destination bit size when checking algebraic optimizations<br></span></blockquote><div><br></div><div>I looked through it today and, by and large, it looks pretty good.  I agree with Connor on Patch 21.  I would like to see patch 13 reworked as per those comments.  I think we should drop patch 23 for now (it needs to go in a series with opt_algebraic changes to use it).<br><br>On the rest of the patches, I had mostly minor comments.  Assuming the comments are addressed, those patches are<br><br></div><div>Reviewed-by: Jason Ekstrand <<a href="mailto:jason@jlekstrand.net">jason@jlekstrand.net</a>><br></div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="">
 src/compiler/Makefile.sources                     |   1 +<br>
 src/compiler/glsl/Makefile.sources                |   1 +<br>
 src/compiler/nir/Makefile.sources                 |   1 +<br>
</span> src/compiler/nir/glsl_to_nir.cpp                  |  89 ++++++++++++------<br>
 src/compiler/nir/nir.c                            |  19 +++-<br>
 src/compiler/nir/nir.h                            |   8 +-<br>
 src/compiler/nir/nir_algebraic.py                 |  11 ++-<br>
 src/compiler/nir/nir_builder.h                    |   4 +-<br>
 src/compiler/nir/nir_clone.c                      |   6 +-<br>
 src/compiler/nir/nir_control_flow.c               |   5 +-<br>
<span class=""> src/compiler/nir/nir_from_ssa.c                   |   2 +<br>
 src/compiler/nir/nir_instr_set.c                  |  22 ++++-<br>
 src/compiler/nir/nir_lower_alu_to_scalar.c        |   3 +<br>
</span> src/compiler/nir/nir_lower_atomics.c              |   5 +-<br>
 src/compiler/nir/nir_lower_double_packing.c       | 106 ++++++++++++++++++++++<br>
 src/compiler/nir/nir_lower_load_const_to_scalar.c |   8 +-<br>
 src/compiler/nir/nir_lower_locals_to_regs.c       |   3 +-<br>
 src/compiler/nir/nir_lower_to_source_mods.c       |  27 +++++-<br>
 src/compiler/nir/nir_lower_vars_to_ssa.c          |   3 +-<br>
 src/compiler/nir/nir_lower_vec_to_movs.c          |   1 +<br>
 src/compiler/nir/nir_opcodes.py                   |  72 +++++++++++++++<br>
 src/compiler/nir/nir_opt_constant_folding.c       |   4 +-<br>
 src/compiler/nir/nir_phi_builder.c                |   3 +-<br>
 src/compiler/nir/nir_print.c                      |  17 +++-<br>
 src/compiler/nir/nir_search.c                     |   5 +-<br>
 src/compiler/nir/nir_split_var_copies.c           |   2 +<br>
 src/compiler/nir/nir_to_ssa.c                     |   2 +-<br>
 src/gallium/auxiliary/nir/tgsi_to_nir.c           |   2 +-<br>
 28 files changed, 366 insertions(+), 66 deletions(-)<br>
<div class="HOEnZb"><div class="h5"> create mode 100644 src/compiler/nir/nir_lower_double_packing.c<br>
<br>
--<br>
2.5.0<br>
<br>
_______________________________________________<br>
mesa-dev mailing list<br>
<a href="mailto:mesa-dev@lists.freedesktop.org">mesa-dev@lists.freedesktop.org</a><br>
<a href="https://lists.freedesktop.org/mailman/listinfo/mesa-dev" rel="noreferrer" target="_blank">https://lists.freedesktop.org/mailman/listinfo/mesa-dev</a><br>
</div></div></blockquote></div><br></div></div>