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.

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

Jump to bottom

New Build System #791

Open
wants to merge 71 commits into
base: main
Choose a base branch
from
Open

New Build System #791

wants to merge 71 commits into from

Conversation

pmocz
Copy link
Member

@pmocz pmocz commented Mar 2, 2025

by Vincent Vanlaer

VincentVanlaer and others added 30 commits January 23, 2025 14:28
@VincentVanlaer
build: add new build system
81d47cd 
This commit introduces a new build system (still based on make files).
Reasons for replacing the previous build system were:

- the previous build system had a lot of duplication between the
  modules, both in terms of make files and shell scripts.
- parallelisation was limited. The new build system parallelises better
  within a module, and between modules.
- build scripts and output from the build ended up in the same folder.
  With the new system, all build output files are stored in the folder
  ``build`` in the repository root, making cleaning build files
  equivalent to deleting this folder

For the details of the implementation, see the newly added documentation
(Developing -> Build system).

Some further notable changes:

- makedepf90 has been replaced by a perl script that scans the fortran
  source files. This allows for better flexibility in construction the
  make files.
- Currently, `pymesa` does not work with the new build system, as the
  new system does not provide a way to build shared libraries for every
  module.
- Dependencies are being discovered using pkg-config. This change
  makes it easier to build MESA without the SDK, as linux distributions
  typically ship with pkg-config files.
@VincentVanlaer
build: support current SDK version
3f31cc7 
This commit adds the necessary pkg-config files until the SDK ships
them.
@VincentVanlaer
const: port to new build system
545e685
@VincentVanlaer
utils: port to new build system
9b7ea0d
@VincentVanlaer
math: port to new build system
6b775ac
@VincentVanlaer
mtx: port to new build system
048df5f
@VincentVanlaer
auto_diff: port to new build system
5f20114
@VincentVanlaer
num: port to new build system
59a8d37
@VincentVanlaer
interp_1d: port to new build system
b4e0a64
@VincentVanlaer
interp_2d: port to new build system
a23ba78
@VincentVanlaer
chem: port to new build system
7898b11
@VincentVanlaer
chem: port preprocessor to new build system
e172320
@VincentVanlaer
eos: port to new build system
f2d69e7
@VincentVanlaer
forum: port to new build system
a668579
@warrickball @VincentVanlaer
colors: port to new build system
dc046cd
@VincentVanlaer
gyre: port to new build system
426c76a
@VincentVanlaer
forum: fix recopying files on build
a2c98a7
@VincentVanlaer
kap: port to new build system
e151878
@VincentVanlaer
kap: port plotter to new build system
c4943c7
@VincentVanlaer
kap: move preprocessing oplib script out of archive
87f56b7 
It makes more sense to place the data processing script under proper
version control.
@VincentVanlaer
kap: port preprocessor to new build system
77dc3df 
Some of the preprocessor code has been modified to actually build, and
to catch some errors in a cleaner way
@VincentVanlaer
kap: cleanup unused files in preprocessor
df1a8c9
@VincentVanlaer
build: add lapack95 pkg-config file
e6f450f 
Needed to build GYRE with the SDK
@VincentVanlaer
gyre: don't combine the different gyre libraries
6229cfb 
This is difficult to do across platforms
@VincentVanlaer
build: replace realpath with equivalent perl implementation
7b6233c
@warrickball @VincentVanlaer
rates: port to new build system
40d87f0
@warrickball @VincentVanlaer
neu: port to new build system
fd477f0
@warrickball @VincentVanlaer
net: port to new build system
3a5df2d
@warrickball @VincentVanlaer
ionization: port to new build system
7a5f57b
@warrickball @VincentVanlaer
atm: port to new build system
117a253
@VincentVanlaer
build: clean up remaining traces of old build system
f2dd2fc
@VincentVanlaer
build: reenable and fix all tests
8cac0e6
@VincentVanlaer
build: do not install modules on a normal build
b409299
@VincentVanlaer
build: clean up install script
2be2e15
@VincentVanlaer
build: remove pushd for portability
18a469f
@VincentVanlaer
build: switch to bash where non-sh features are used
28b80bb
@VincentVanlaer
fixup! star: port dev cases to new build system
af05250
@VincentVanlaer
fixup! build: clean up install script
45aa18b
@VincentVanlaer
shmesa: small changes for new build system
36bdc61
@VincentVanlaer
fixup! binary: port test suite to new build system
f170331
@VincentVanlaer
fixup! astero: port test suite to new build system
d22a8b4
@VincentVanlaer
fixup! build: reenable and fix all tests
5225be5
@VincentVanlaer
build: make it possible to disable FPE traps for compatibility with e…
cd773f9 
…arlier versions of MESA
@VincentVanlaer
docs: update for new build system
8baecdd
@VincentVanlaer
fixup! build: clean up install script
03efdb1
@VincentVanlaer
fixup! build: clean up install script
fecf573
@warrickball
Copy link
Contributor

I decided that I'll fight with ifort after the build system is refreshed, so I configured our University cluster to run using the SDK. It submitted test results over the course of last night. There are 25 failures but I want to get some big picture input on that result.

tl;dr: We have two options:

  1. We could probably go back to whatever compiler options meant that stuff runs under the current build system.
  2. We can try to fix a lot of issues that have been raised first.

I'm strongly in favour of 2. I think these runs have flagged some genuine issues that are worth fixing now rather than sweeping back under a rug of compiler options. Or maybe some of these are fixed on main or just missing one or two reasonable compiler flags?

Regarding those issues, they actually fall into some large categories.

  • double_bh, 15M_dynamo, accreted_material_j, high_rot_darkening and ppisn fail because of a checksum difference in the final model when it's computed from a restart. This is presumably a symptom of something not being correctly passed into and out of the photo, though I can't tell at a glance if it's the same thing.
  • 1M_thermohaline, 20M_z2m2_high_rotation, conserve_angular_momentum and magnetic_braking hit a shared issue, where it looks like s% D_ST(2) is NaN on restart. I guess this is also something not being restored from a photo correctly.
  • The RSP tests all fail because of the same energy conservation test.
  • The astero failures are all caused by index out of bounds errors in ADIPLS. I'll make investigating these my next priority.

The remaining tests (carbon_kh, relax_composition_j_entropy and simplex_solar_calibration) have unique issues. simplex_solar_calibration has an unusual structure that includes a bit of astero that BillP isolated. I think the build system just needs updating here.

@warrickball
Copy link
Contributor

Also probably worth mentioning that this is a continuation of this original external PR (#724).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@warrickball warrickball Awaiting requested review from warrickball warrickball is a code owner

@orlox orlox Awaiting requested review from orlox orlox is a code owner

@mjoyceGR mjoyceGR Awaiting requested review from mjoyceGR mjoyceGR is a code owner

@Debraheem Debraheem Awaiting requested review from Debraheem Debraheem is a code owner

@aurimontem aurimontem Awaiting requested review from aurimontem aurimontem is a code owner

@wmwolf wmwolf Awaiting requested review from wmwolf wmwolf is a code owner

@evbauer evbauer Awaiting requested review from evbauer evbauer is a code owner

@fxt44 fxt44 Awaiting requested review from fxt44 fxt44 is a code owner

@annethoul annethoul Awaiting requested review from annethoul annethoul is a code owner

@rsmolec rsmolec Awaiting requested review from rsmolec rsmolec is a code owner

@aarondotter aarondotter Awaiting requested review from aarondotter aarondotter is a code owner

@rhdtownsend rhdtownsend Awaiting requested review from rhdtownsend rhdtownsend is a code owner

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@pmocz @warrickball @VincentVanlaer