With current master we're seeing build failures with AOSP: error: undefined symbol: ir3_nir_lower_imul

This is due to the ir3_nir_imul.c file not being generated in the Android.mk files.

This patch simply adds it to the Android build, after which thigns build and boot ok on db410c.

Cc: Rob Clark robdclark@chromium.org Cc: Emil Velikov emil.l.velikov@gmail.com Cc: Amit Pundir amit.pundir@linaro.org Cc: Sumit Semwal sumit.semwal@linaro.org Cc: Alistair Strachan astrachan@google.com Cc: Greg Hartman ghartman@google.com Cc: Tapani Pälli tapani.palli@intel.com Signed-off-by: John Stultz john.stultz@linaro.org --- src/freedreno/Makefile.sources | 3 ++- src/gallium/drivers/freedreno/Android.gen.mk | 7 +++++++ 2 files changed, 9 insertions(+), 1 deletion(-)

diff --git a/src/freedreno/Makefile.sources b/src/freedreno/Makefile.sources index d8aaf2caecc..75ec361663b 100644 --- a/src/freedreno/Makefile.sources +++ b/src/freedreno/Makefile.sources @@ -48,7 +48,8 @@ ir3_SOURCES := \ ir3/ir3_sun.c

ir3_GENERATED_FILES := \ - ir3/ir3_nir_trig.c + ir3/ir3_nir_trig.c \ + ir3/ir3_nir_imul.c

registers_FILES := \ registers/a2xx.xml.h \ diff --git a/src/gallium/drivers/freedreno/Android.gen.mk b/src/gallium/drivers/freedreno/Android.gen.mk index d29ba159d5c..1d3ee5ff856 100644 --- a/src/gallium/drivers/freedreno/Android.gen.mk +++ b/src/gallium/drivers/freedreno/Android.gen.mk @@ -28,11 +28,18 @@ ir3_nir_trig_deps := \ $(MESA_TOP)/src/freedreno/ir3/ir3_nir_trig.py \ $(MESA_TOP)/src/compiler/nir/nir_algebraic.py

+ir3_nir_imul_deps := \ + $(MESA_TOP)/src/freedreno/ir3/ir3_nir_imul.py + intermediates := $(call local-generated-sources-dir)

$(intermediates)/ir3/ir3_nir_trig.c: $(ir3_nir_trig_deps) @mkdir -p $(dir $@) $(hide) $(MESA_PYTHON2) $< -p $(MESA_TOP)/src/compiler/nir > $@

+$(intermediates)/ir3/ir3_nir_imul.c: $(ir3_nir_imul_deps) + @mkdir -p $(dir $@) + $(hide) $(MESA_PYTHON2) $< -p $(MESA_TOP)/src/compiler/nir > $@ + LOCAL_GENERATED_SOURCES += $(addprefix $(intermediates)/, \ $(ir3_GENERATED_FILES))

Thanks for the notice. I had already had a fix for this, but forgot to push it. It's pushed now.

Marek

On Wed, Jul 3, 2019 at 7:10 PM John Stultz john.stultz@linaro.org wrote:

Building with clang, I'm seeing u_blitter.h:627:1: error: control may reach end of non-void function [-Werror,-Wreturn-type]

The util_blitter_get_color_format_for_zs() asserts for any unhandled types, so we do not expect to reach the end of the function here.

But provide a dummy return with an explicit assert above to ensure we don't hit it with any future changes to the logic.

Cc: Rob Clark robdclark@chromium.org Cc: Emil Velikov emil.l.velikov@gmail.com Cc: Amit Pundir amit.pundir@linaro.org Cc: Sumit Semwal sumit.semwal@linaro.org Cc: Alistair Strachan astrachan@google.com Cc: Greg Hartman ghartman@google.com Cc: Tapani Pälli tapani.palli@intel.com Cc: Marek Olšák marek.olsak@amd.com Signed-off-by: John Stultz john.stultz@linaro.org

---

src/gallium/auxiliary/util/u_blitter.h | 3 +++ 1 file changed, 3 insertions(+)

diff --git a/src/gallium/auxiliary/util/u_blitter.h b/src/gallium/auxiliary/util/u_blitter.h index 9e3fa55e648..7d6c3db64da 100644 --- a/src/gallium/auxiliary/util/u_blitter.h +++ b/src/gallium/auxiliary/util/u_blitter.h @@ -624,6 +624,9 @@ util_blitter_get_color_format_for_zs(enum pipe_format format) default: assert(0); }

• assert(0);
• /*XXX NEVER GET HERE*/
• return PIPE_FORMAT_R32G32_UINT;

}

#ifdef __cplusplus

2.17.1

---

dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel