This is the mail archive of the
binutils@sourceware.cygnus.com
mailing list for the binutils project.
Re: ld test failure on ARM GNU/Linux
- To: alan at linuxcare dot com dot au
- Subject: Re: ld test failure on ARM GNU/Linux
- From: Ian Lance Taylor <ian at zembu dot com>
- Date: 10 Apr 2000 14:32:15 -0700
- CC: Philip dot Blundell at pobox dot com, binutils at sourceware dot cygnus dot com, rth at cygnus dot com
- References: <Pine.LNX.4.21.0004101328310.4828-100000@front.linuxcare.com.au>
Date: Mon, 10 Apr 2000 13:55:45 +1000 (EST)
From: Alan Modra <alan@linuxcare.com.au>
The question boils down to: Are the comments correct in the
following? If so, then ld/testsuite/ld-selective/selective.exp is in
error. The test fails due to finding foo__1B, and the comments seem to
indicate that foo_1B should indeed be kept.
I dunno. How smart is the gc code? How smart is gcc?
With the 2.95.2 x86 compiler I am using, gcc is smart enough to
generate a vtable_entry for A::foo and to not generate one for B::foo.
Given that, the code may be correct in discarding B::foo. I'm not
completely sure, though. I don't know where the precise semantics of
the VTENTRY and VTINHERIT relocations are defined, if anywhere.
Alan, with what compiler version do you see this test fail on the x86?
Philip, what compiler version are you testing with? Richard, you
wrote this code in the first place; what would you expect to happen
with the ld/testsuite/ld-selective/3.cc code?
Ian