Skip to content

Fix ARM half conversion in RenderOutputWriter #36

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
gripeyes wants to merge 1 commit into
base: main
Choose a base branch
from
+9 −8
Open

Fix ARM half conversion in RenderOutputWriter #36

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 9 additions & 8 deletions lib/rendering/rndr/RenderOutputWriter.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1568,10 +1568,10 @@ float
RenderOutputWriter::htof(const unsigned short h)
{

#if defined(__ARM_NEON__) // TODO: Verify this
float output;
vst1q_f32(&output, vcvt_f32_f16(vld1_u16(&h)));
return output;
#if defined(__ARM_NEON__)
__fp16 input;
std::memcpy(static_cast<void *>(&input), static_cast<const void *>(&h), sizeof(input));
return static_cast<float>(input);
#else
return _cvtsh_ss(h); // Convert half 16bit float to full 32bit float
#endif
Expand All @@ -1583,10 +1583,11 @@ unsigned short
RenderOutputWriter::ftoh(const float f)
{

#if defined(__ARM_NEON__) // TODO: Verify this
__fp16 output;
vst1_f16(&output, vcvt_f16_f32(vld1q_f32(&f)));
return output;
#if defined(__ARM_NEON__)
__fp16 h = static_cast<__fp16>(f);
unsigned short output;
std::memcpy(static_cast<void *>(&output), static_cast<const void *>(&h), sizeof(output));
return output;
#else
return _cvtss_sh(f, 0); // Convert full 32bit float to half 16bit float
// An immediate value controlling rounding using bits : 0=Nearest
Expand Down