projects
/
ghc-hetmet.git
/ commitdiff
commit
grep
author
committer
pickaxe
?
search:
re
summary
|
shortlog
|
log
|
commit
| commitdiff |
tree
raw
|
patch
|
inline
| side by side (parent:
ae8fc4d
)
Added a comment about se.info.type being used uninitialized
author
sven.panne@aedion.de
<unknown>
Fri, 10 Nov 2006 16:26:54 +0000
(16:26 +0000)
committer
sven.panne@aedion.de
<unknown>
Fri, 10 Nov 2006 16:26:54 +0000
(16:26 +0000)
rts/RetainerProfile.c
patch
|
blob
|
history
diff --git
a/rts/RetainerProfile.c
b/rts/RetainerProfile.c
index
cd00013
..
17f5afb
100644
(file)
--- a/
rts/RetainerProfile.c
+++ b/
rts/RetainerProfile.c
@@
-662,6
+662,12
@@
push( StgClosure *c, retainer c_child_r, StgClosure **first_child )
// following statement by either a memcpy() call or a switch statement
// on the type of the element. Currently, the size of stackElement is
// small enough (5 words) that this direct assignment seems to be enough.
// following statement by either a memcpy() call or a switch statement
// on the type of the element. Currently, the size of stackElement is
// small enough (5 words) that this direct assignment seems to be enough.
+
+ // ToDo: The line below leads to the warning:
+ // warning: 'se.info.type' may be used uninitialized in this function
+ // This is caused by the fact that there are execution paths through the
+ // large switch statement above where some cases do not initialize this
+ // field. Is this really harmless? Can we avoid the warning?
*stackTop = se;
#ifdef DEBUG_RETAINER
*stackTop = se;
#ifdef DEBUG_RETAINER