Mercurial > emacs
changeset 111368:5a8705defdf9
Backport r99751 from trunk
| author | Chong Yidong <cyd@stupidchicken.com> |
|---|---|
| date | Thu, 04 Nov 2010 15:54:14 -0400 |
| parents | 1d1b5e1c230b |
| children | e7a9eb3aa4ac |
| files | src/ChangeLog src/process.c |
| diffstat | 2 files changed, 5 insertions(+), 14 deletions(-) [+] |
line wrap: on
line diff
--- a/src/ChangeLog Thu Nov 04 15:53:28 2010 -0400 +++ b/src/ChangeLog Thu Nov 04 15:54:14 2010 -0400 @@ -1,3 +1,8 @@ +2010-03-25 YAMAMOTO Mitsuharu <mituharu@math.s.chiba-u.ac.jp> + + * process.c (Fmake_network_process): Don't call turn_on_atimers around + `connect' (Bug#5723). + 2010-03-25 Helmut Eller <eller.helmut@gmail.com> * process.c (Fmake_network_process): Call `select' for interrupted
--- a/src/process.c Thu Nov 04 15:53:28 2010 -0400 +++ b/src/process.c Thu Nov 04 15:54:14 2010 -0400 @@ -3654,23 +3654,9 @@ immediate_quit = 1; QUIT; - /* This turns off all alarm-based interrupts; the - bind_polling_period call above doesn't always turn all the - short-interval ones off, especially if interrupt_input is - set. - - It'd be nice to be able to control the connect timeout - though. Would non-blocking connect calls be portable? - - This used to be conditioned by HAVE_GETADDRINFO. Why? */ - - turn_on_atimers (0); - ret = connect (s, lres->ai_addr, lres->ai_addrlen); xerrno = errno; - turn_on_atimers (1); - if (ret == 0 || xerrno == EISCONN) { /* The unwind-protect will be discarded afterwards.