[libc] -Wshorten-64-to-32 in libc/src/__support/File/linux/file.cpp for arm32

[nickdesaulniers]

https://lab.llvm.org/buildbot/#/builders/229/builds/22038

/llvm/libc_worker/worker/libc-arm32-debian/libc-arm32-debian-dbg/llvm-project/libc/src/__support/File/linux/file.cpp:50:17: error: implicit conversion loses integer precision: 'long long' to 'long' [-Werror,-Wshorten-64-to-32]
  return result.value();
         ~~~~~~~^~~~~~~

looks like another issue with off_t vs long for 32b ARM

[llvmbot]

@llvm/issue-subscribers-libc

Author: Nick Desaulniers (nickdesaulniers)

https://lab.llvm.org/buildbot/#/builders/229/builds/22038

/llvm/libc_worker/worker/libc-arm32-debian/libc-arm32-debian-dbg/llvm-project/libc/src/__support/File/linux/file.cpp:50:17: error: implicit conversion loses integer precision: 'long long' to 'long' [-Werror,-Wshorten-64-to-32]
  return result.value();
         ~~~~~~~^~~~~~~

looks like another issue with off_t vs long for 32b ARM

[nickdesaulniers]

I think this was fixed by #77810.