RFR: JDK-8034788 Rewrite toolchain.m4 to support multiple toolchains per platform
henry.jen at oracle.com
Thu Feb 13 02:33:21 UTC 2014
Good progress, it build for me on Arch Linux with gcc, not working for
clang yet(perhaps expected). A few comments from what I read so far,
- --with-toolchain-type=list shows up after quite a few configure going
on already, we probably want to have this done in an earlier stage.
- The "available" toolchains feels to me should be available on the
build system, i.e., they are installed and available to use. From what
is implemented, it's more "valid choices" than "available".
- On the above note, should "list" list "available" or "valid choices"?
- Do we want to allow convention name cc and CC and detect what it is or
simply ask for explicit choice as it is imeplemted now? Although I don't
think it's likely for a system to install, say gcc, as cc instead of
create a symbolic link alias for gcc, I cannot be sure.
- I understand clang is not a goal for this patch. From experience, it's
almost work like gcc. Does it make sense to have gcc flags as a
fallback? This bring up the question before, should we have a fallback
default to unknown toolchain? Even if we check for supported toolchain,
all such setting should probably still be defensive with gcc as default.
Otherwise, I can configure with clang but quite a few setting would be
- On windows, gcc is listed as an available toolchain, I think this is
good. However, we still do TOOLCHAIN_SETUP_VISUAL_STUDIO_ENV if select
gcc as toolchain. That should not be checked for gcc?
On 02/12/2014 06:58 AM, Magnus Ihse Bursie wrote:
> We need to be able to handle different toolchains on the same platform,
> e.g. both gcc and clang on macosx. We have felt the need for this for
> quite some time, but it has become pressing due to the clang on macosx
> issue. (We are on a path to replace gcc with clang as official build
> platform on macosx, but we need to handle both toolchains during a
> transition period.
> To do this properly requires a complete reorganisation and cleanup of
> This is, as you can see, quite a large change in the configure system.
> Unfortunately, due to the bad state of the code in toolchain.m4, it was
> hard to do it in smaller steps. I have still tried to keep the changes
> small. For instance:
> * Full clang functionality was not a goal for this patch.
> * Fixing all issues in toolchain.m4 was not a goal for this patch.
> (Although I've sprinkled the code with FIXME that I intend to address as
> soon as possible.)
> Both of these will be addressed in follow-up patches.
> In this step, the priority has been to:
> 1) introduce the toolchain type concept,
> 2) cleanup toolchain.m4 to be able support this, and
> 3) move flag handling out of toolchain.m4
> This is a (somewhat cleaned up) log from the commits to my internal
> working repo, it gives a hint on what kind of changes/refactorisations
> that were performed:
> * Separate flag handling from toolchain.m4 into new file flags.m4.
> * Remove all non-flags handling from flags code (this has aldready been
> handled elsewhere).
> * Clean up flags.m4 heavily. Rename functions FLAGS_*. Group setting of
> flags by flag type. Simplify logic if possible. Adapt to new toolchain
> * Adapt configure.ac to how stuff have moved and splitted.
> * Remove old COMPILER_TYPE and COMPILER_NAME, fully replace with
> * Print toolchain and improved compiler version info in
> * Move AC_SUBST to where variables were set up.
> * Move TOOLCHAIN_SETUP_MSVCR_DLL etc into new function
> LIB_SETUP_ON_WINDOWS in libraries.m4.
> * Fix compiler version output for microsoft compiler.
> * Fix version parsing for solstudio.
> * Reorder and rename functions in toolchain.m4.
> * Move strip and mcs flags setup to flags.m4.
> * Break out Windows compiler platform check from
> * Make sure --with-tools-dir override everything, even on windows.
> * Add/update documentation for all functions in toolchain.m4
> * Add/improve implementation comments.
> * Add lots of FIXME:s for stuff needing fixes (but not this time around).
> * Remove AS_IF
> * Fix broken indentation.
> * Fix bug in printing in BASIC_FIXUP_EXECUTABLE.
> Bug: https://bugs.openjdk.java.net/browse/JDK-8034788
More information about the build-dev