[GHC] #9395: Debug.Trace should not use %s for format string

[GHC]

#9395: Debug.Trace should not use %s for format string
-------------------------------------+-------------------------------------
              Reporter:  ezyang      |            Owner:  ezyang
                  Type:  bug         |           Status:  new
              Priority:  normal      |        Milestone:
             Component:              |          Version:  7.8.2
  libraries/base                     |         Keywords:
            Resolution:              |     Architecture:  Unknown/Multiple
      Operating System:              |       Difficulty:  Unknown
  Unknown/Multiple                   |       Blocked By:
       Type of failure:  Incorrect   |  Related Tickets:
  result at runtime                  |
             Test Case:              |
              Blocking:              |
Differential Revisions:              |
-------------------------------------+-------------------------------------

Comment (by ezyang):

 OK, it's a little trickier than just specifying a new format string; I
 forgot that

 {{{
 #include "stdio.h"
 int main() {
     printf("%.*s", 4, "AA\0B");
     return 0;
 }
 }}}

 only outputs AA, even with the length description. So we can't use
 debugBelch. This raises the question, why are we using debugBelch at all?
 Some archaeology later:

 {{{
 [project @ 2005-01-28 23:33:57 by krasimir]
 - The output from uncaught exceptions handler is redirected to RTS's
 errorBelch.
 - The output from Debug.Trace is redirected to RTS's debugBelch
 - Usually errorBelch and debugBelch messages go to stderr except for
 Windows GUI applications. For GUI applications the Debug.Trace output is
 redirected to debug console and the exceptions message is displayed in
 message box.
 }}}

 It also seems like a good idea generally to bypass Haskell-lands
 buffering. Unfortunately, printf style output is a hard-coded assumption
 for `RtsMsgFunction`, so we can't easily swap it out without introducing
 another function for raw output.

 Maybe, alternately, Debug.Trace should munge out null bytes?

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9395#comment:2>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler