This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Dwarf2out end prologue/begin epilogue for IVMS (2nd try)


Richard Henderson wrote:
On 09/09/2009 04:04 PM, Douglas B Rupp wrote:
The rewrite started to get very ugly, so I took a step back and added a
flag arg to ASM_OUTPUT_DWARF_DELTA instead.

What do you think?

I don't like that at all. I preferred you using the new dw_val_class entry (dw_cal_class_vms_delta?) which neatly isolates the change away from ASM_OUTPUT_DWARF_DELTA.


+      fprintf (file, "\tdata4.ua\t (");
+      assemble_name (file, lab1);
+      fprintf (file, "-");
+      assemble_name (file, lab2);
+      fprintf (file, ")/16*3");

Given that lab1 and lab2 are marked as debugging labels, they aren't necessarily on bundle boundaries. I'm not sure that this construction works reliably.


Are you using gas, or the system assembler? If gas, I'd think about adding some sort of directive that can actually count slots for you here, like the ia64 unwind directives do.

Gas. Will investigate that route.



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]