Closed barracuda156 closed 1 month ago
On Friday September 06 2024 07:40:20 Sergey Fedorov wrote:
Perhaps an upstream bug? I can open an issue with them.
Looks like it, though I see I updated port:nspr just after port:nss and since no ABI was broken I didn't try to build nss anew.
You could try the current version of port:nss, or merge the current one from MacPorts to see if that solves anything.
FWIW, it seems that the specificity of these two ports in LinuxPorts is that nspr is installed to $prefix/lib and not in a subdir. I can't remember what reason I had to do that, possibly just to avoid having to add another rpath component to all dependents.
@RJVB There are two similar reported cases: https://bugzilla.mozilla.org/show_bug.cgi?id=483135 https://bugzilla.mozilla.org/show_bug.cgi?id=747371
From that it seems something might have built for 32-bit, god knows why. I can check that tomorrow.
On Friday September 06 2024 07:40:20 Sergey Fedorov wrote:
drbg.c:615:5: error: 'arg' declared as an array with a negative size PR_STATIC_ASSERT(sizeof(size_t) <= 4); ^
~~~~~~~~ /opt/local/include/nspr/prtypes.h:556:42: note: expanded from macro 'PR_STATIC_ASSERT' extern void pr_static_assert(int arg[(condition) ? 1 : -1]) ^~~~~~~~Perhaps an upstream bug? I can open an issue with them.
This in fact looks like a requirement that the size_t type be no more than 4 bytes long. Which is probably not appropriate for an OS called pine64 running on a riscv64 platform.
Perhaps an upstream bug? I can open an issue with them.