https://sourceware.org/bugzilla/show_bug.cgi?id=20475
--- Comment #10 from whitequark at whitequark dot org ---
Unfortunately the problem resists my attempts to write a minimal testcase. I do
not have any more time to spend debugging this, although I will gladly test any
suggestions.
As a fix, I
https://sourceware.org/bugzilla/show_bug.cgi?id=20475
whitequark at whitequark dot org changed:
What|Removed |Added
Status|RESOLVED|UNCONFIRMED
https://sourceware.org/bugzilla/show_bug.cgi?id=20475
--- Comment #6 from whitequark at whitequark dot org ---
I've verified that your suggested fix works. However, I don't have commit bit
for binutils.
--
You are receiving this mail because:
You are on the CC list f
https://sourceware.org/bugzilla/show_bug.cgi?id=20475
--- Comment #4 from whitequark at whitequark dot org ---
Created attachment 9442
--> https://sourceware.org/bugzilla/attachment.cgi?id=9442&action=edit
OR1K target patch
--
You are receiving this mail because:
You are on the CC l
https://sourceware.org/bugzilla/show_bug.cgi?id=20475
--- Comment #3 from whitequark at whitequark dot org ---
Actually, I might be wrong about this, since the comment in elflink.c implies
that the behavior is intentional. I'll attach another patch, which fixes the
logic of reloc
https://sourceware.org/bugzilla/show_bug.cgi?id=20475
--- Comment #1 from whitequark at whitequark dot org ---
The previous comment is not quite right.
1) The _GLOBAL_OFFSET_TABLE_ did not point at the end of .got.plt but at the
start.
2) This is a bug in the architecture-independent ELF backend
https://sourceware.org/bugzilla/show_bug.cgi?id=20475
--- Comment #2 from whitequark at whitequark dot org ---
Created attachment 9439
--> https://sourceware.org/bugzilla/attachment.cgi?id=9439&action=edit
Patch
--
You are receiving this mail because:
You are on the CC list for
Version: 2.25
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: ld
Assignee: unassigned at sourceware dot org
Reporter: whitequark at whitequark dot org
Target Milestone: ---
I am using ld 2.25.1 built for or1k-elf target
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #20 from whitequark at whitequark dot org ---
Thanks!
--
You are receiving this mail because:
You are on the CC list for the bug.
___
bug-binutils mailing list
bug-binutils@gnu.org
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #17 from whitequark at whitequark dot org ---
I use --target=or1k-linux, since I need to generate shared object files. I
still run them on bare metal but it looks like -linux is the closest I have.
--
You are receiving this mail
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #15 from whitequark at whitequark dot org ---
Ah right, sorry. I messed up the git invocation somehow, it seems.
--
You are receiving this mail because:
You are on the CC list for the bug
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #13 from whitequark at whitequark dot org ---
Nick, that can't be right. I've just looked at binutils gitweb (e.g.
https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=blob;f=bfd/elf3
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
whitequark at whitequark dot org changed:
What|Removed |Added
Attachment #8474|0 |1
is
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #10 from whitequark at whitequark dot org ---
Ok, I took a shot at fixing gas today. To recap...
Here is the minimal testcase:
.section .data1
.long 0
x: .long 0
.section .data2
.long 0
.long 0
y: .long (x
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #9 from whitequark at whitequark dot org ---
I have confirmed that, at least, gas emits PC-relative relocations incorrectly.
I will fix that.
Speaking of the ABI: OR1K does not have a document (any document) which
specifies the
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #7 from whitequark at whitequark dot org ---
Why did you revert this?
--
You are receiving this mail because:
You are on the CC list for the bug.
___
bug-binutils mailing list
bug
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #5 from whitequark at whitequark dot org ---
I have looked at this and it is not a bug I've introduced. The test xfails
or32, which is how or1k used to be named. The test was not updated, which has
in fact hidden this very bug
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
--- Comment #4 from whitequark at whitequark dot org ---
I will take a look at this soon.
--
You are receiving this mail because:
You are on the CC list for the bug.
___
bug-binutils mailing list
https://sourceware.org/bugzilla/show_bug.cgi?id=18759
whitequark at whitequark dot org changed:
What|Removed |Added
CC||whitequark at
: binutils
Assignee: unassigned at sourceware dot org
Reporter: whitequark at whitequark dot org
Target Milestone: ---
Created attachment 8474
--> https://sourceware.org/bugzilla/attachment.cgi?id=8474&action=edit
Patch
Currently, the OpenRISC1000 relocations R_OR1K_{
20 matches
Mail list logo
