Bwa-mem2: fix compilation issue

[apraga]

On latest unstable, there are several implicit function declarations in safestrlib, which are new errors instead of warnings.
safestrlib was not a seperate package but simply a submodule in this nix packages.

I have separated it into a library and updated it to latest unstable, which essentially fixes the issue.

Also, there are no tests upstream (the test folder is for benchmarking).

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandboxing enabled in nix.conf? (See Nix manual)
  • sandbox = relaxed
  • sandbox = true
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 25.05 Release Notes (or backporting 24.11 and 25.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
• Fits CONTRIBUTING.md.

---

Add a 👍 reaction to pull requests you find important.

[apraga]

@alxsimon What do you think of this separation ? I'm not 100% sure it compiles all arch as before though.
@GaetanLepage for testing and your help in general :)
Thanks

[alxsimon]

@alxsimon What do you think of this separation ? I'm not 100% sure it compiles all arch as before though. @GaetanLepage for testing and your help in general :) Thanks

Hi, I don't have enough experience with this to have an opinion sorry.
As I'm not using Nix anymore, my name should probably be removed from the maintainer field of this package, feel free to replace mine with yours in this PR for example, or a following one.

[GaetanLepage]

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 380725

---

x86_64-linux

✅ 2 packages built:

• bwa-mem2
• safestringlib

---

aarch64-linux

✅ 1 package built:

• safestringlib

---

x86_64-darwin

❌ 2 packages failed to build:

• bwa-mem2
• safestringlib

---

aarch64-darwin

❌ 1 package failed to build:

• safestringlib

[GaetanLepage]

Fails on darwin with:

ld: unknown option: -z
clang: error: linker command failed with exit code 1 (use -v to see invocation)
make[2]: *** [CMakeFiles/safestring_shared.dir/build.make:226: libsafestring_shared.1.2.0.dylib] Error 1
make[1]: *** [CMakeFiles/Makefile2:147: CMakeFiles/safestring_shared.dir/all] Error 2

[apraga]

We are in luck : a fix has been proposed upstream. The patch was added : does it compile on darwin now ?

[GaetanLepage]

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 380725

---

x86_64-linux

✅ 2 packages built:

• bwa-mem2
• safestringlib

---

aarch64-linux

✅ 1 package built:

• safestringlib

---

x86_64-darwin

❌ 2 packages failed to build:

• bwa-mem2
• safestringlib

---

aarch64-darwin

❌ 1 package failed to build:

• safestringlib

[GaetanLepage]

safestrings still fails with:

/tmp/nix-build-safestringlib-1.2.0-unstable-2024-10-21.drv-1/source/unittests/test_memset_s.c:30:35: error: too few arguments to function call, expected 4, have 3
   30 |     rc = memset_s(NULL, LEN, value);
      |          ~~~~~~~~

[apraga]

Yes, the tests weren't there initially.... Here's a patch (untested).

[GaetanLepage]

Still failing...

/tmp/nix-build-safestringlib-1.2.0-unstable-2024-10-21.drv-1/source/unittests/test_wcscpy_s.c:289:5: error: call to undeclared function 'wmemset_s'; ISO C99 and later do not support implicit function declarations [-Wimplicit-function-declaration]
  289 |     wmemset_s(str1, L'x', 20);
      |     ^

[apraga]

Okay... I've tried to disable this test specifically. Let me know how it goes...

[GaetanLepage]

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 380725

---

x86_64-linux

✅ 2 packages built:

• bwa-mem2
• safestringlib

---

aarch64-linux

✅ 1 package built:

• safestringlib

---

x86_64-darwin

❌ 2 packages failed to build:

• bwa-mem2
• safestringlib

---

aarch64-darwin

❌ 1 package failed to build:

• safestringlib

[apraga]

Is it the same error as before ?

[GaetanLepage]

It is now failing at the very end:

----------------END TEST--------------------
Running phase: installPhase
cp: cannot stat '../libsafestring_shared.so': No such file or directory

[apraga]

Could you check if libsafestring_shared.so exits on darwin ? If it's just libsafestring_shared.a that is built, or if the name is different, we can manage. If nothing is built, we'll have to mark it as broken I guess. Sorry :/

[GaetanLepage]

Could you check if libsafestring_shared.so exits on darwin ?

Here is what the folder contains at the beginning of the installPhase:

----------------END TEST--------------------
Running phase: installPhase
---------------------------------------
total 356
drwxr-xr-x 13 _nixbld1 nixbld    416 Feb 13 00:00 .
drwxr-xr-x 21 _nixbld1 nixbld    672 Feb 13 00:00 ..
-rw-r--r--  1 _nixbld1 nixbld  25479 Feb 13 00:00 CMakeCache.txt
drwxr-xr-x 16 _nixbld1 nixbld    512 Feb 13 00:00 CMakeFiles
-r--r--r--  1 _nixbld1 nixbld   4346 Feb 13 00:00 CPackConfig.cmake
-r--r--r--  1 _nixbld1 nixbld   4841 Feb 13 00:00 CPackSourceConfig.cmake
-rw-r--r--  1 _nixbld1 nixbld  83647 Feb 13 00:00 Makefile
-rw-r--r--  1 _nixbld1 nixbld   8730 Feb 13 00:00 cmake_install.cmake
-rwxr-xr-x  1 _nixbld1 nixbld  87840 Feb 13 00:00 libsafestring_shared.1.2.0.dylib
lrwxr-xr-x  1 _nixbld1 nixbld     32 Feb 13 00:00 libsafestring_shared.dylib -> libsafestring_shared.1.2.0.dylib
-rw-r--r--  1 _nixbld1 nixbld 125000 Feb 13 00:00 libsafestring_static.a
-rw-r--r--  1 _nixbld1 nixbld   2796 Feb 13 00:00 safestringConfig.cmake
drwxr-xr-x  6 _nixbld1 nixbld    192 Feb 13 00:00 unittests
cp: cannot stat '../libsafestring_shared.so': No such file or directory

[apraga]

I've removed the .so alltogether as it's not needed by bwa-mem2. Does that fixes it ?

[GaetanLepage]

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 380725

---

x86_64-linux

✅ 2 packages built:

• bwa-mem2
• safestringlib

---

aarch64-linux

✅ 1 package built:

• safestringlib

---

x86_64-darwin

❌ 1 package failed to build:

• bwa-mem2

✅ 1 package built:

• safestringlib

---

aarch64-darwin

✅ 1 package built:

• safestringlib

[GaetanLepage]

safestringlib now builds fine on all platforms.
bwa-mem2 fails on x86_64-darwin:

clang++ -g -O3 -fpermissive -msse -msse2 -msse3 -mssse3 -msse4.1   src/main.o libbwa.a -lpthread -lm -lz -L. -lbwa -L/nix/store/qxi18xq5pncpb28n46pgg7xx4isc7ff9-safestringlib-1.2.0->
ld: library not found for -l:libsafestring.a
clang++: error: linker command failed with exit code 1 (use -v to see invocation)
make: *** [Makefile:107: bwa-mem2] Error 1

[apraga]

Almost there ! Could you check libsafestring.a is indeed in $out/lib for safestring ?

[GaetanLepage]

Almost there ! Could you check libsafestring.a is indeed in $out/lib for safestring ?

nixpkgs on  bwa-mem2-fix
❯ nix-build -A safestringlib --system x86_64-darwin
/nix/store/qxi18xq5pncpb28n46pgg7xx4isc7ff9-safestringlib-1.2.0-unstable-2024-10-21

nixpkgs on  bwa-mem2-fix
❯ ls -l result/lib/
.r--r--r-- 111k root  1 Jan  1970 libsafestring.a

[apraga]

I've fixed the linker flag. It should work with clang. I tried to test it by overriding stdenv but it failed on my machine so I defer to you :)

[GaetanLepage]

nixpkgs-review result

Generated using nixpkgs-review.

Command: nixpkgs-review pr 380725

---

x86_64-linux

✅ 2 packages built:

• bwa-mem2
• safestringlib

---

aarch64-linux

✅ 1 package built:

• safestringlib

---

x86_64-darwin

✅ 2 packages built:

• bwa-mem2
• safestringlib

---

aarch64-darwin

✅ 1 package built:

• safestringlib

[GaetanLepage]

Good job @apraga !