Opened 10 years ago
Closed 10 years ago
#3497 closed task (fixed)
inconsistency concerning -fomit-frame-pointer in gcc builds
| Reported by: | Joel Kammet | Owned by: | |
|---|---|---|---|
| Priority: | normal | Milestone: | 7.5 |
| Component: | Book | Version: | SVN |
| Severity: | normal | Keywords: | |
| Cc: |
Description
This issue was previously mentioned in ticket #2363 but doesn't seem to have been addressed.
In section 5.10 the explanation refers to "x86 machines" and the sed patch is written unconditionally so it is applied equally to 32 and 64 bit platforms.
Section 6.17 says "As in Section 5.10 ... apply the following" but the sed command is written in a case statement "case uname -m in i?86)" so it is not applied to 64 bit machines (where uname -m` returns "x86_64").
If this inconsistency is intended it should be explained. And if it is not intended, how is it that it hasn't caused problems in all these years?
Change History (3)
comment:1 by , 10 years ago
comment:3 by , 10 years ago
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
Note:
See TracTickets
for help on using tickets.
This goes a long way back. The original sed was introduced in September 2005 before we supported x86_64. At that time both Chapter 5 and Chapter 6 were the same. The change to Chapter 6 was made in December 2008.
This was brought up due to an issue identified in the iterative change analysis (ica) that is no longer done due to technical difficulties.
This issue was brought up on the -dev mailing list in December 2010. In that discussion, the issue was not completely resolved. I'm not sure the omit_frame_pointer change is needed at all any more, but I'll change Chapter 5 to be identical with the Chapter 6 construct.