Re: Bug in test-fcntl.c

2021-05-10 Thread Nicholas Gaya
The process hierarchy starts with tclsh (MacPorts), then a bunch of layers of make and sh (from oath-toolkit's recursive `make check`), and finally the test-fcntl executable. I was using MacPorts' "trace mode", which seems to be responsible for the particularly high number of inherited open file

Re: maint.mk suggestion for gnupload

2021-05-10 Thread Eric Blake
On 5/8/21 6:34 AM, Eric Blake wrote: > but it would be a lot nicer if I didn't have to remember to do that > after the fact. In fact, reading gnupload --help says it is as simple > as calling: > > $ gnupload --to ... --symlink-regex ... > > in the first place. But --symlink-regex is not listed

Re: getrandom: Fail with ENOSYS when the system has no randomness source

2021-05-10 Thread Paul Eggert
On 5/9/21 7:51 AM, Bruno Haible wrote: if (fd < 0) -return fd; +{ + errno = ENOSYS; + return -1; +} If 'open' fails with errno equal to (say) EINTR or EAGAIN or EMFILE, this doesn't mean the operating system lacks a randomness source; it merel

Re: ptrdiff_t overflow checks for malloc-posix etc.

2021-05-10 Thread Paul Eggert
Thanks for all those fixes (and the detailed reports!).