[llvm-bugs] [Bug 48201] New: [LLVM 11 regression] Linker check, new in LLVM 11, rejects common GCC code
via llvm-bugs
llvm-bugs at lists.llvm.org
Tue Nov 17 04:46:16 PST 2020
https://bugs.llvm.org/show_bug.cgi?id=48201
Bug ID: 48201
Summary: [LLVM 11 regression] Linker check, new in LLVM 11,
rejects common GCC code
Product: libraries
Version: 11.0
Hardware: All
OS: Linux
Status: NEW
Keywords: compile-fail, regression
Severity: normal
Priority: P
Component: Linker
Assignee: unassignedbugs at nondot.org
Reporter: burnus at net-b.de
CC: llvm-bugs at lists.llvm.org
LLVM's linker added a reasonable check in https://reviews.llvm.org/D73999
Such as (from the testsuite):
.section .foo,"aM", at progbits,1
.section .foo,"aM", at progbits,4
error: changed section entsize for .foo, expected: 1
However, it also now causes an error for:
.section .rodata.cst8,"aM", at progbits,8
.section .rodata.cst8
The problem is that GCC generates this by default (gcc/varasm.c):
/* If we have already declared this section, we can use an
abbreviated form to switch back to it -- unless this section is
part of a COMDAT groups, in which case GAS requires the full
declaration every time. */
if (!(HAVE_COMDAT_GROUP && (flags & SECTION_LINKONCE))
&& (flags & SECTION_DECLARED))
{
fprintf (asm_out_file, "\t.section\t%s\n", name);
return;
}
LLVM and GNU as ("gas") tried to align, cf. discussion
https://sourceware.org/legacy-ml/binutils/2020-02/msg00091.html
and the GAS patch to turn a warning to an error:
https://sourceware.org/legacy-ml/binutils/2020-02/msg00129.html
However, while llvm/lib/MC/MCParser/ELFAsmParser.cpp always issues an error:
+ if (Section->getType() != Type)
+ Error(loc, "changed section type for " + SectionName + ", expected: 0x" +
+ utohexstr(Section->getType()));
etc.
The GNU assembler only does so if flags have been specified,
gas/config/obj-elf.c:
if (attr != 0)
{
/* If section attributes are specified the second time we see a
particular section, then check that they are the same as we
saw the first time. */
if (((old_sec->flags ^ flags)
...
--
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20201117/dcb89d29/attachment.html>
More information about the llvm-bugs
mailing list