Fix gen_context/ASM build on ARM #936

pull real-or-random wants to merge 3 commits into bitcoin-core:master from real-or-random:202105-gen-context-no-ext-asm changing 2 files +7 −2
  1. real-or-random commented at 1:11 PM on May 3, 2021: contributor

    Obsoletes #935.

  2. gen_context: Don't use external ASM because it complicates the build
    Fixes #931.
    99f47c20ec
  3. Makefile.am: Honor config when building gen_context
    This passes $(DEFS) (which should literally be "-DHAVE_CONFIG_H") to the
    compiler when building gen_context.
    
    This has currently no effect because gen_context.c does not check for
    this macro but it's conceivable that it may do so in the future.
    2161f31785
  4. Makefile.am: Don't pass a variable twice c8483520c9
  5. gmaxwell commented at 1:39 PM on May 3, 2021: contributor

    I will test now.

  6. gmaxwell commented at 1:49 PM on May 3, 2021: contributor

    ACK c8483520c9077905a1dc8b9adb88b6ea2a3bd9ef looks good and works here. Undefign is kinda yuck, but it is already doing it and it's cleaner than the obvious alternatives.

  7. sipa commented at 7:14 PM on May 3, 2021: contributor

    utACK c8483520c9077905a1dc8b9adb88b6ea2a3bd9ef. I verified that building still works on ARM64, but without asm of course.

    Mental note for something to add: there is no sanity checking in ./configure when --with-asm=arm is selected, so it "succeeds" on ARM64, but compilation fails of course.

  8. real-or-random commented at 7:28 PM on May 3, 2021: contributor

    Mental note for something to add: there is no sanity checking in ./configure when --with-asm=arm is selected, so it "succeeds" on ARM64, but compilation fails of course.

    In generally, this makes sense because auto is always the clever setting and setting a value explicitly just forces this value. Maybe we should rename simply rename the setting to arm32?

  9. sipa commented at 7:43 PM on May 3, 2021: contributor

    @real-or-random Well, I think even when forcing we can check whether the build environment we have supports that option (we do for x86_64). Renaming to arm32 makes sense.

  10. sipa merged this on May 4, 2021
  11. sipa closed this on May 4, 2021

  12. surinder83singh referenced this in commit 3855601992 on Jun 2, 2021
  13. vmta referenced this in commit 780707485f on Jun 9, 2021
  14. jonasnick referenced this in commit 6db00f5b2e on Jul 13, 2021
  15. sipa referenced this in commit c020cbaa5c on Jul 14, 2021
  16. apoelstra referenced this in commit 3fcc6fcacd on Jul 28, 2021
  17. apoelstra referenced this in commit 196c993d1f on Jul 28, 2021
  18. furszy referenced this in commit 534b2b2f5a on Sep 23, 2021
  19. janus referenced this in commit 62075fc9aa on Nov 5, 2021
  20. patricklodder referenced this in commit 21badcf9d2 on Jul 25, 2022
  21. patricklodder referenced this in commit 03002a9013 on Jul 28, 2022
  22. str4d referenced this in commit ed3e5b0f2a on Sep 27, 2022
  23. Fabcien referenced this in commit c084621a20 on Jan 15, 2026
  24. Fabcien referenced this in commit d61aab1152 on Jan 15, 2026

github-metadata-mirror

This is a metadata mirror of the GitHub repository bitcoin-core/secp256k1. This site is not affiliated with GitHub. Content is generated from a GitHub metadata backup.
generated: 2026-05-19 06:52 UTC