large repo clone issue

[shuheque]

Hello,

I have a large repository which is more than 30GB, when I clone it with git-for-windows (tried 2.47.1 and some old versions), it prompted me with "fatal: pack has bad object at offset xxxxx: inflate returned 1", usually died at "receiving objects" stage at about 4GiB.

Looks like it's same issue in #2179 ; Seems like it's a "long" versus "size_t" problem and should be fixed long ago, but why it's still there? I was using 64bit git and 64bit windows server;

Thanks!!

[dscho]

The Pull Request is still open because it has not been fixed. That Pull Request is in a terrible, unreviewable shape.

Can you use WSL's git instead? That is the only work-around I am currently aware of.

[dscho]

@shuheque You can sure try to build! However, you may run into serious trouble if you try to merge (or rebase to) a newer Git version. My comment from way back when still applies: It is too big of a patch and needs to be broken down. And it can be broken down, as I have described. I even demonstrated how that will totally work.

[shuheque]

@dscho Thanks Yeah, I tried to build your fork of branch "size_t6", but seems there is only a perl generator instead of CMakes under contrib/buildsystems, this perl can generate projects but seems can't set INCLUDE dirs for libs under VCPKG, such as libconv.

Did I miss anything?
VCPKG has been booted with the bat script already and those packages are in place.

[shuheque]

@dscho Ohh, I just make MSVC=1 and fixed a bunch of library issue and it's working, also need to do this with GIT SDK. Thanks!!

[dscho]

For the record, I would recommend rebasing to the current main (and resolving all the merge conflicts, which might take a while) and building Git by following this documentation.

Actually, it might be easier not even to bother with rebasing but instead following the idea to convert some callers from unsigned long to size_t, incrementally building up to a size_t-only call stack.

[shuheque]

For the record, I would recommend rebasing to the current main (and resolving all the merge conflicts, which might take a while) and building Git by following this documentation.

Actually, it might be easier not even to bother with rebasing but instead following the idea to convert some callers from unsigned long to size_t, incrementally building up to a size_t-only call stack.

Yep, and this is what's been done at my side, your size_t6 branch works just fine, thanks!!