CALL_INSNs and LIBCALLs

Richard Henderson rth@cygnus.com
Fri Apr 14 16:04:00 GMT 2000


On Fri, Apr 14, 2000 at 06:06:17PM -0400, Richard Kenner wrote:
> Some CALL_INSNs end a basic block, but they must not do so if they are
> in a LIBCALL.  This fixes that.

This is wrong.  They should have already not been ending a 
basic block due to an REG_EH_REGION note being -1.


r~


More information about the Gcc-patches mailing list