#4954: -eventlog / -debug should imply -rtsopts
---------------------------------+------------------------------------------
    Reporter:  duncan            |       Owner:              
        Type:  feature request   |      Status:  new         
    Priority:  normal            |   Component:  Driver      
     Version:  7.0.1             |    Keywords:              
    Testcase:                    |   Blockedby:              
          Os:  Unknown/Multiple  |    Blocking:              
Architecture:  Unknown/Multiple  |     Failure:  None/Unknown
---------------------------------+------------------------------------------
 The `-eventlog` flag links the program to the threaded rts. The eventlog
 is only generated when using the `-l` RTS flag, but by default we get
 {{{
 $ ./foo +RTS -l
 Setup: Most RTS options are disabled. Link with -rtsopts to enable them.
 }}}
 You have to use both `-eventlog` and `-rtsopts` to be able to run with
 `+RTS -l`.

 I think `-eventlog` should imply `-rtsopts` because it's pointless to use
 the `-eventlog` way if one never uses the runtime flag `+RTS -l`.

 The `-rtsopts` was added as a security measure, so that by default the
 runtime `+RTS -RTS` options would not be available. Since `-eventlog` is a
 non-default option then I think it is also safe from a security POV to
 have it imply `-rtsopts`.

 For the security paranoid, at most, one could consider in the userguide
 where `-rtsopts` is mentioned and the security issue pointed out, that it
 could also mention what other flags imply `-rtsopts`.


 BTW, I tried making a patch for this, but the `-eventlog` flag is a static
 flag while `-rtsopts` is a dynamic flag and it appears hard to make one
 imply the other (the only case where that is done is labeled as a hack).

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/4954>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler

_______________________________________________
Glasgow-haskell-bugs mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs

Reply via email to