Skip to content
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.

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

development: Move target specific code to a single subdirectory. #624

Closed
wants to merge 2 commits into from

Conversation

jbowler
Copy link
Contributor

@jbowler jbowler commented Oct 14, 2024

This is a two commit move of all the current target specific code to a single
subdirectory called 'target'.

  • development: Move target specific code to a subdirectory.
  • development: Move target specific code to a subdirectory.

This commit only contains the "move" part, additional changes are
necessary to the build system to support the new locations.

Signed-off-by: John Bowler <[email protected]>
This completes the change to move all target specific code to the
subdirectory 'target' by making the appropriate changes to the three
build systems.

Signed-off-by: John Bowler <[email protected]>
Copy link
Member

@ctruta ctruta left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I was looking forward to get this integrated, next thing after the release of 1.6.45 (which I had planned for early this week), but stuff came out and I couldn't publish 1.6.45 yet.

After release, I will resync libpng18 with libpng16 through a "tag merge", to reduce (at least for a little bit) the divergence between the two branches, but the code should largely stay the same.

However, before we make a change that will be then frozen for the next decade or so, let us work a little bit on the nomenclature, because I'm not so thrilled about the new directory name target.

The concept of "target" makes a lot of sense in CMake, and a lot less sense in Autotools and (unfortunately) in Meson which (unfortunately) followed up on Autotools' nomenclature. It makes less sense in source code. On top of that, "target" means "target platform" which means "machine" plus "operating system". Our code in question is machine-specific; or, rather, architecture-specific (or arch-specific); or, rather, SIMD-specific. Seriously, that code is about nothing else but SIMD optimizations (not about CPU stuff that is not SIMD stuff, and also, not about target platform stuff that is not CPU stuff).

MOREOVER:

As I noticed how various libpng users are vendoring our library, they tend to only include the code that is part of libpng, and not projects, and not scripts, and not contrib, etc. In order to make it easier for everybody to know what is libpng vs. what is NOT libpng, I propose using the prefix png for everything in the top-level directory that is indeed libpng.

IN CONCLUSION:

I propose that we move all of the machine-specific stuff (or arch-specific stuff, or SIMD-specific stuff) to a directory named pngmachine (or pngarch, or pngsimd). My personal favourites are pngarch and pngsimd, but I'm open to a debate.

AND ALSO:

I propose that we move all the pnglibconf stuff from the scripts directory to a more glorified pnglibconf directory, so that we can easily explain to everybody that all top-level png*/ directories (i.e. (pngsimd/ or pngarch/ or pngmachine/) and (pnglibconf/)) should always travel with all the top-level png* files (i.e. png*.c and png*.h).

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

filter_neon.S was a dummy placeholder in 1.6.x, which we kept in, to avoid breaking stuff. This commit is the opportunity to delete it for good.

@jbowler jbowler closed this Oct 18, 2024
@jbowler jbowler deleted the tsc-source-mv branch October 18, 2024 21:55
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants