Summary: | Configure options are poorly documented | ||
---|---|---|---|
Product: | gcc | Reporter: | Paolo Carlini <paolo.carlini> |
Component: | libstdc++ | Assignee: | Not yet assigned to anyone <unassigned> |
Status: | NEW --- | ||
Severity: | normal | CC: | bkoz, egallager, gcc-bugs, iains, pinskia, sjames, webrown.cpp |
Priority: | P2 | Keywords: | documentation |
Version: | 4.0.0 | ||
Target Milestone: | --- | ||
See Also: |
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=13756 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=82824 |
||
Host: | Target: | ||
Build: | Known to work: | ||
Known to fail: | Last reconfirmed: | 2005-12-15 04:52:33 |
Description
Paolo Carlini
2005-05-13 14:48:05 UTC
Confirmed, --with-pic being documented would be very nice even though it is a generic libtool configure option, it is most used with libstdc++ as shown by the bug reports about the using static libstdc++ in a shared library. Benjamin, any opinion about this issue? Would love to see something standardized for all libs in gcc. My preference is to have libstdc++ config docs linked to in gcc docs. (No duplication, just alias.) One thing that could be done would be to add a "C++ Specific Option" to http://gcc.gnu.org/install/configure.html And link to http://gcc.gnu.org/onlinedocs/libstdc++/manual/configure.html Or options on that page, don't care. As an aside, are a lot of these java config opts actually still relevant? -benjamin There's always ./configure --help=recursive from the top level; the "recursive" makes it print the help text from subdirs like libstdc++ too. The issue of undocumented configure options also came up in bug 82824, although that was a tangent so I won't add it under "See Also" (In reply to Benjamin Kosnik from comment #3) > > As an aside, are a lot of these java config opts actually still relevant? > > -benjamin No, java has been removed. (In reply to Eric Gallager from comment #5) > The issue of undocumented configure options also came up in bug 82824, > although that was a tangent so I won't add it under "See Also" > Actually I've changed my mind; I now believe that even tangents are sufficient for "See Also" status. (In reply to Andrew Pinski from comment #1) > Confirmed, --with-pic being documented would be very nice even though it is > a generic libtool > configure option, it is most used with libstdc++ as shown by the bug reports > about the using static > libstdc++ in a shared library. Note that documenting it has the risk of leading people to think that --with-pic (or --without-pic) is ok to use even in situations when it's not; e.g. x86_64 must always be PIC. If the flag is going to be documented, any such restrictions should be documented as well. retitling |