mirror of
git://gcc.gnu.org/git/gcc.git
synced 2024-12-22 23:09:54 +08:00
b2dad0e372
2000-04-21 Benjamin Kosnik <bkoz@redhat.com> * libstdc++-v3: New directory. From-SVN: r33317
148 lines
3.9 KiB
HTML
148 lines
3.9 KiB
HTML
|
|
<!--#include virtual="/libstdc++/header.html"-->
|
|
|
|
|
|
<H3>Status</H3>
|
|
<P>Compiling this library requires a recent version of gcc. (For instance,
|
|
gcc 2-95 works well.) In addition, <A HREF="/autoconf/">autoconf</A>
|
|
versions 2.13 and above and <A HREF="/automake/">automake</A> versions
|
|
1.3b and above are required and essential ingredients for those mucking
|
|
about in the Makefiles and trying to configure on new targets.
|
|
</P>
|
|
|
|
<P>New things that work are in the latest
|
|
<A HREF="17_intro/RELEASE-NOTES">release notes</A>.
|
|
</P>
|
|
|
|
<P>Known problems and bugs, which will force us to actually keep this
|
|
file up-to-date, can be found in <A HREF="17_intro/BUGS">bugs</A>.
|
|
</P>
|
|
|
|
<P>Things that need to be done and are looking for volunteers can be found
|
|
in: <A HREF="17_intro/TODO">pick me</A>.
|
|
</P>
|
|
|
|
<P>This release has been tested with gcc 2.95 on the following
|
|
platforms. The configuration of --disable-namespaces
|
|
--disable-libgcc-rebuild was tested. Results for the
|
|
--enable-namespaces builds should be comparable. Here are the
|
|
results for snapshots 2.90.N:
|
|
</P>
|
|
|
|
<TABLE CELLSPACING=4 CELLPADDING=1>
|
|
|
|
<!-- Column titles -->
|
|
<TR>
|
|
<TH ALIGN=LEFT>.N</TH>
|
|
<TH ALIGN=LEFT>host</TH>
|
|
<TH ALIGN=CENTER>build</TH>
|
|
<TH ALIGN=CENTER>static pass/fail</TH>
|
|
<TH ALIGN=CENTER>shared pass/fail</TH>
|
|
</TR>
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>alpha-unknown-linux-gnu</TD>
|
|
<TD ALIGN=CENTER>+ (see Note 1)</TD>
|
|
<TD ALIGN=CENTER>62/5</TD>
|
|
<TD ALIGN=CENTER>62/5</TD>
|
|
</TR>
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>i686-unknown-linux-gnu</TD>
|
|
<TD ALIGN=CENTER>+</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
</TR>
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>powerpc-unknown-linux-gnu</TD>
|
|
<TD ALIGN=CENTER>+</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
</TR>
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>sparc-sun-solaris2.5.1</TD>
|
|
<TD ALIGN=CENTER>+ (see Note 2)</TD>
|
|
<TD ALIGN=CENTER>62/6</TD>
|
|
<TD ALIGN=CENTER>62/6</TD>
|
|
</TR>
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>sparc-sun-solaris2.6</TD>
|
|
<TD ALIGN=CENTER>+ (see Note 2)</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
</TR>
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>sparc-sun-solaris2.7</TD>
|
|
<TD ALIGN=CENTER>+ (see Note 2)</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
</TR>
|
|
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>powerpc-ibm-aix4.x</TD>
|
|
<TD ALIGN=CENTER>- (see Note 3)</TD>
|
|
<TD ALIGN=CENTER>0/68</TD>
|
|
<TD ALIGN=CENTER>0/68</TD>
|
|
</TR>
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>i686-cygwin</TD>
|
|
<TD ALIGN=CENTER>+</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
<TD ALIGN=CENTER>0/68</TD>
|
|
</TR>
|
|
|
|
<TR>
|
|
<TD ALIGN=LEFT>.8</TD>
|
|
<TD ALIGN=LEFT>i686-linux-x-powerpc-elf</TD>
|
|
<TD ALIGN=CENTER>+</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
<TD ALIGN=CENTER>65/3</TD>
|
|
</TR>
|
|
|
|
</TABLE>
|
|
|
|
<P>Note 1: A newer compiler than gcc-2.95.2 is required.
|
|
</P>
|
|
|
|
<P>Note 2: All solaris toolchains were configured with --with-gnu-ld
|
|
--with-gnu-as --with-gnu-binutils --with-gcc. (That's a lot of --with's...)
|
|
</P>
|
|
|
|
<P>Note 3: Configuration and initial support for AIX has been
|
|
completed. However, problems with the compiler/linker have
|
|
prevented testing. For instance, mknumeric_limits will fail unless
|
|
weak linking is possible, which doesn't seem possible on
|
|
AIX. The initial bug report is
|
|
<a href="http://gcc.gnu.org/ml/gcc-bugs/2000-03/msg00574.html"> here </a>
|
|
|
|
<p>And the
|
|
response from one of the powerpc port maintainers:
|
|
|
|
<p>
|
|
<i>
|
|
I opposed the original solution (involving weak symbols) on ELF, as
|
|
it seemed to be the wrong idea. I'm not sure what the right fix is,
|
|
and don't think I have time to work on it right now.</i>
|
|
</p>
|
|
</P>
|
|
|
|
|
|
|
|
|
|
<!--#include virtual="/libstdc++/footer.html"-->
|
|
|