Opened 7 years ago

Closed 7 years ago

#3654 closed task (fixed)

Some redundancies in toolchain checks

Reported by: chris@… Owned by: lfs-book@…
Priority: normal Milestone: 7.6
Component: Book Version: SVN
Severity: normal Keywords:


First, most of the sentences mentioning what the output should be start with "If everything is working correctly, there should be no errors[...]" and I think there's no need to keep repeating that. It's also inaccurate anyway, since those are all just grep commands, so of course there wouldn't be any errors, just lack of output at worst.

Second, I don't see why the dynamic linker location is checked twice. First a readelf command is run on the compiled test program, then the last command in the series of checks is to look for the linker name in the log - is there any situation where these commands would give different output?

Change History (3)

comment:1 by bdubbs@…, 7 years ago

This ticket is not completely clear. Are you *only* referring to 6.10. Adjusting the Toolchain?

Second part:

I think the two different checks for the dynamic linker are checking two slightly different things. The first is asking what linker is specifying in a.out. The second is whether that linker is actually found on the correct path by ld. Admittedly it would be difficult for one to pass and the other to fail. However, I don't see a problem with leaving the tests as is.

comment:2 by chris@…, 7 years ago

Yes, sorry, this is referring to the toolchain checks in Chapter 6, both in Adjust the Toolchain and the GCC page (though most of them are the same anyway, with xincludes).

comment:3 by bdubbs@…, 7 years ago

Resolution: fixed
Status: newclosed

Fixed at revision 10693.

Note: See TracTickets for help on using tickets.