Re: [Gcl-devel] Re: gcl_signal

[Michael Koehne]

Moin Mike,

> I'll leave the ins and outs of that one to you and Camm!

  the signal code is something I dont like to touch before reading its
  implications a few more times. But sigfpe1 & sigpipe smell ugly :

  sigfpe1 is not delayed by main_signal_handler, but running directly,
  calling FEerror - breaking in garbage collector state ...
  sigpipe is calling perror(""); and FEerror next - the empty string
  for perror really stinks for half patched/half broken ...

  so I suggest to add both signals to signals_handled, and to remove
  the perror("") from sigpipe. I think, i can construct a test condition
  for a sigpipe and user installed sigpipe, lets look how it works out.

Bye Michael
-- 
  mailto:address@hidden             UNA:+.? 'CED+2+:::Linux:2.4.22'UNZ+1'
  http://www.xml-edifact.org/           CETERUM CENSEO WINDOWS ESSE DELENDAM