Is the intent to revert the change of type to the second argument of
PERL_HASH?
If it is not then there is a cast needed at line 590 of vms/vms.c where the
"show logical *" output is used (I think to initialize %ENV).
I do have a patch that helps the VMS build avoid breaking on:
Copy/Log/Noconfirm [.vms]vms.c []
%COPY-S-COPIED, DKA100:[PPRYMMER.PERL.VMS]VMS.C;1 copied to
DKA100:[PPRYMMER.PERL]VMS.C;1 (462 blocks)
CC/DECC/NOANSI_ALIAS /Include=[]/Standard=Relaxed_ANSI/Prefix=All
/Obj=.obj/NoList/float=ieee/ieee=denorm_results/Define=PERL_CORE VMS.C
PERL_HASH(hash,key,keylen);
.......^
%CC-W-PTRMISMATCH1, In the initializer for s_PeRlHaSh, the referenced type
of
the pointer value "key" is "char", which is not compatible with
"const unsigned char" because they differ by signed/unsigned attribute.
at line number 590 in file DKA100:[PPRYMMER.PERL]VMS.C;1
%MMS-F-ABORT, For target VMS.OBJ, CLI returned abort status: %X10B91260.
Which way is this headed? Thanks.
Peter Prymmer
