Runs: 715684864 Best found: 00000001d6e8e98fb21368b96a075ae30c6c8e39 (31/28 bits) Time: 289.670476 ~2.470686MH/s000000

I tried 32 bits first but stopped after it took too long. Then I tried 28 and it found 31 zero bits rather quickly. But it did not stop and use that hash and instead continued searching for 28.

I would have been satisfied with 31 of course.

It would be nice if you could at least choose to accept longer hashes.

I think it might even deadlock if it finds more bits than required. Hence why it hangs for finding small amounts of zero bits (1,2,3 even 10)

With cmake -B build -DOPENSSL_ROOT_PATH=/usr/local/opt/openssl, I am getting below error:

CMake Error at /usr/local/Cellar/cmake/3.20.5/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:230 (message):
  Could NOT find OpenSSL, try to set the path to OpenSSL root folder in the
  system variable OPENSSL_ROOT_DIR (missing: OPENSSL_CRYPTO_LIBRARY
  OPENSSL_INCLUDE_DIR)

With cmake -B build -DOPENSSL_ROOT_DIR=/usr/local/opt/openssl, its working fine. I am using cmake version 3.20.5

Using Git CLI like this:

git log --show-signature

Right now it shows a message like gpg: unknown armor header: Nonce: UXS^O AAAAA1316 (please see the screenshot).

MR fixes this, you can pull my commit to compare it against yours locally.
I'm not sure the fix is clearly correct, but it works also for other commits.

Even if you choose not to, here's the fixes for anyone trying with cmake 3.13.4 and libgit2 0.27.7:

CMake Error at CMakeLists.txt:29 (install):
  install TARGETS given no RUNTIME DESTINATION for executable target
  "git-power".
...
git-power/main.cpp: In function ‘void try_commits(size_t, size_t, size_t, char*, size_t)’:
git-power/main.cpp:300:47: error: ‘GIT_OBJECT_COMMIT’ was not declared in this scope
  git_odb_hash(&hash, commit_buf, commit_size, GIT_OBJECT_COMMIT);
                                               ^~~~~~~~~~~~~~~~~

diff --git i/CMakeLists.txt w/CMakeLists.txt
index 4604d57..f5a26d6 100644
--- i/CMakeLists.txt
+++ w/CMakeLists.txt
@@ -26,4 +26,4 @@ if (UNIX AND NOT APPLE)
     target_link_libraries(git-power PRIVATE pthread atomic)
 endif ()
 
-install(TARGETS git-power)
+install(TARGETS git-power RUNTIME DESTINATION bin)
diff --git i/main.cpp w/main.cpp
index 4cf9a9e..983bc9d 100644
--- i/main.cpp
+++ w/main.cpp
@@ -297,7 +297,7 @@ I should not be allowed near this
 
        // Big sanity check here since we think this is a good hash
        git_oid hash;
-       git_odb_hash(&hash, commit_buf, commit_size, GIT_OBJECT_COMMIT);
+       git_odb_hash(&hash, commit_buf, commit_size, GIT_OBJ_COMMIT);
        int test = 0;
        for (size_t n = 0; n < bits; n++) {
                // Test will be the OR of the first `n` bits
@@ -473,7 +473,7 @@ int main(int argc, const char **argv) {
 
                // And make the commit for them
                git_oid hash;
-               git_odb_write(&hash, db, commit_buffer, commit_size, GIT_OBJECT_COMMIT);
+               git_odb_write(&hash, db, commit_buffer, commit_size, GIT_OBJ_COMMIT);
 
                // Fancy print
                char id[0x100];
@@ -486,7 +486,7 @@ int main(int argc, const char **argv) {
 
                // Soft reset to this commit so it is now branch head
                git_object *new_commit;
-               git_object_lookup(&new_commit, repository, &hash, GIT_OBJECT_COMMIT);
+               git_object_lookup(&new_commit, repository, &hash, GIT_OBJ_COMMIT);
 
                git_reset(repository, new_commit, GIT_RESET_SOFT, nullptr);
 

• Added a proper command line option parsing utility (CLI11 https://github.com/CLIUtils/CLI11)
• Updated the headers to be the proper ones for C++
• Moved C headers into the proper C++ extern section.
• Changed current directory to "./" which should be understood by any modern operating system.
• Added optional compiler flags for better debugging.

And I might have tashed your formatting by accident.

Since it seems like more people have joined the "customize git commit hashes because why not" bandwagon (welcome!), I figured I'd share some insights from doing a lot of performance tuning on my own implementation from a few years ago. Obviously, none of this should be taken to discourage new implementations or to ruin any fun.

In decreasing order of effectiveness, the main things that helped were:

• Running on a GPU (this improved performance by roughly a factor of 10, depending on hardware)
• Caching SHA1 buffer state. The tl;dr is that if you put the nonce at the end of your commit message (or signature, if applicable), you can cache the state of the SHA1 algorithm as applied to all but the last ~64 bytes of the commit, and then reapply the cached state to the last 64 bytes on each attempt. This improved performance by a factor of about 4-5 or more, depending on how big the commit is.
• If the nonce is always the same length, then the trailer that the SHA1 puts at the end of data when hashing is also effectively fixed, and I was able to improve the performance of lucky-commit by about 25% by not recomputing it each time.

(Also mentioning @mkrasnitski from https://github.com/mkrasnitski/git-power-rs)