[LLVMdev] Bug in Cloning.h?
Mario Schwalbe
mario at se.inf.tu-dresden.de
Sun Apr 29 08:37:18 PDT 2012
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Hi all,
unless I overlook something, the second version of CloneFunction()
(from http://llvm.org/doxygen/Cloning_8h_source.html):
00114 Function *CloneFunction(const Function *F,
00115 ValueToValueMapTy &VMap,
00116 bool ModuleLevelChanges,
00117 ClonedCodeInfo *CodeInfo = 0);
00118
00119 /// CloneFunction - Version of the function that doesn't need the VMap.
00120 ///
00121 inline Function *CloneFunction(const Function *F, ClonedCodeInfo *CodeInfo = 0){
00122 ValueToValueMapTy VMap;
00123 return CloneFunction(F, VMap, CodeInfo);
00124 }
seems to pass the pointer argument CodeInfo as boolean ModuleLevelChanges
to the first one never filling CodeInfo with data.
Sincerely,
Mario
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iQEcBAEBAgAGBQJPnWAoAAoJEDv0fP6GapNtJUAH/3b/NDdAPE6Dm5YtSkwfl1d9
tIeYp/gfc64MqrvQM0Vv4SVZFEKbCASGqMVL+vTBCHHAiamDJ7KmoN5NNyqaaU4e
BnPBrbXD1vH4Ckbq9C2Ey6xzVOIX89pvtd5IS00nDfCC+sn2KtPUkk9fXKv9Kjz2
6OWqRLaCpYVVl1PSt9dA2xIxZm13Trc1v9j20o32M00/j79fcCUm7aKKg7oCHhq6
oKt+k+uGR4+4cHVjx5QOjkPJ4PFAbZk4kDHmg06SR94zZJ+mIrKJ+R2pH2DWML2J
JW9887ln/yH5Bh58NT6fUK7AdkXKq0fY8Rt20f4Wn6hOXfhqlwz7Lx03EG5ppLI=
=CzXq
-----END PGP SIGNATURE-----
More information about the llvm-dev
mailing list