This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: "mvfc" fails regression for binutils-2.20.51
Hi Suma,
I am afraid there is another issue that I am facing in the LD regression this
time. The test case "rgn-over8" is failing with the following message:
regexp "^ 1 .data 0+0000400 0+0001000 0+0000400 [0-9a-f]+ 2\*\*[0-9]+ CONTENTS, ALLOC, LOAD, DATA$"
line " 1 .data 00000400 00001000 00001000 00002000 2**0 CONTENTS, ALLOC, LOAD, DATA"
This took some tracking down. It turns out that there is a problem with
the Renesas loader. It ignores the VMA field in the program header and
only uses the LMA field. Thus the RX port of the linker has to set the
LMA field to the VMA value, so that code will execute at the address
that it is expecting. (The Renesas loader does not support loading
code to a different address and then moving it before execution).
Hence this particular test, and the others like it, fail.
I am going to check in the attached patch to mark these tests as
expected failures for the RX target.
Cheers
Nick
ld/testsuite/ChangeLog
2009-11-18 Nick Clifton <nickc@redhat.com>
* ld-elf/extract-symbol-1sec.d: Expect to fail for the RX target.
* ld-scripts/overlay-size.d: Likewise.
* ld-scripts/rgn-at1.d: Likewise.
* ld-scripts/rgn-at2.d: Likewise.
* ld-scripts/rgn-at3.d: Likewise.
* ld-scripts/rgn-at4.d: Likewise.
* ld-scripts/rgn-over8-ok.d: Likewise.
Index: ld/testsuite/ld-elf/extract-symbol-1sec.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-elf/extract-symbol-1sec.d,v
retrieving revision 1.4
diff -c -3 -p -r1.4 extract-symbol-1sec.d
*** ld/testsuite/ld-elf/extract-symbol-1sec.d 29 Sep 2008 14:12:02 -0000 1.4
--- ld/testsuite/ld-elf/extract-symbol-1sec.d 18 Nov 2009 16:35:25 -0000
***************
*** 3,9 ****
#ld: -Textract-symbol-1.ld
#objcopy_linked_file: --extract-symbol
#objdump: --headers
! #xfail: "hppa*-*-*"
#...
Sections:
*Idx +Name +Size +VMA +LMA .*
--- 3,10 ----
#ld: -Textract-symbol-1.ld
#objcopy_linked_file: --extract-symbol
#objdump: --headers
! #xfail: "hppa*-*-*" "rx-*-*"
! # FAILS on the RX because the linker has to set LMA == VMA for the Renesas loader.
#...
Sections:
*Idx +Name +Size +VMA +LMA .*
Index: ld/testsuite/ld-scripts/overlay-size.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-scripts/overlay-size.d,v
retrieving revision 1.3
diff -c -3 -p -r1.3 overlay-size.d
*** ld/testsuite/ld-scripts/overlay-size.d 14 Oct 2004 12:54:47 -0000 1.3
--- ld/testsuite/ld-scripts/overlay-size.d 18 Nov 2009 16:35:26 -0000
***************
*** 1,6 ****
--- 1,9 ----
# ld: -T overlay-size.t -Map tmpdir/overlay-size.map
# name: overlay size
# objdump: --headers
+ # xfail: rx-*-*
+ # FAILS on the RX because the linker has to set LMA == VMA for the
+ # Renesas loader.
#
# The .bss[123] LMAs are deliberately blanked out. We can't
# reliably map overlaid sections to segments.
Index: ld/testsuite/ld-scripts/rgn-at1.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-scripts/rgn-at1.d,v
retrieving revision 1.2
diff -c -3 -p -r1.2 rgn-at1.d
*** ld/testsuite/ld-scripts/rgn-at1.d 22 Apr 2009 11:57:35 -0000 1.2
--- ld/testsuite/ld-scripts/rgn-at1.d 18 Nov 2009 16:35:26 -0000
***************
*** 2,7 ****
--- 2,10 ----
# source: rgn-at.s
# ld: -T rgn-at1.t
# objdump: -w -h
+ # xfail: rx-*-*
+ # FAILS on the RX because the linker has to set LMA == VMA for the
+ # Renesas loader.
.*: file format .*
Index: ld/testsuite/ld-scripts/rgn-at2.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-scripts/rgn-at2.d,v
retrieving revision 1.2
diff -c -3 -p -r1.2 rgn-at2.d
*** ld/testsuite/ld-scripts/rgn-at2.d 22 Apr 2009 11:57:35 -0000 1.2
--- ld/testsuite/ld-scripts/rgn-at2.d 18 Nov 2009 16:35:26 -0000
***************
*** 2,7 ****
--- 2,10 ----
# source: rgn-at.s
# ld: -T rgn-at2.t
# objdump: -w -h
+ # xfail: rx-*-*
+ # FAILS on the RX because the linker has to set LMA == VMA for the
+ # Renesas loader.
.*: file format .*
Index: ld/testsuite/ld-scripts/rgn-at3.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-scripts/rgn-at3.d,v
retrieving revision 1.2
diff -c -3 -p -r1.2 rgn-at3.d
*** ld/testsuite/ld-scripts/rgn-at3.d 22 Apr 2009 11:57:35 -0000 1.2
--- ld/testsuite/ld-scripts/rgn-at3.d 18 Nov 2009 16:35:26 -0000
***************
*** 2,7 ****
--- 2,10 ----
# source: rgn-at.s
# ld: -T rgn-at3.t
# objdump: -w -h
+ # xfail: rx-*-*
+ # FAILS on the RX because the linker has to set LMA == VMA for the
+ # Renesas loader.
.*: file format .*
Index: ld/testsuite/ld-scripts/rgn-at4.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-scripts/rgn-at4.d,v
retrieving revision 1.1
diff -c -3 -p -r1.1 rgn-at4.d
*** ld/testsuite/ld-scripts/rgn-at4.d 22 May 2009 11:27:44 -0000 1.1
--- ld/testsuite/ld-scripts/rgn-at4.d 18 Nov 2009 16:35:26 -0000
***************
*** 2,7 ****
--- 2,10 ----
# source: rgn-at.s
# ld: -T rgn-at4.t
# objdump: -w -h
+ # xfail: rx-*-*
+ # FAILS on the RX because the linker has to set LMA == VMA for the
+ # Renesas loader.
.*: file format .*
Index: ld/testsuite/ld-scripts/rgn-over8-ok.d
===================================================================
RCS file: /cvs/src/src/ld/testsuite/ld-scripts/rgn-over8-ok.d,v
retrieving revision 1.2
diff -c -3 -p -r1.2 rgn-over8-ok.d
*** ld/testsuite/ld-scripts/rgn-over8-ok.d 2 May 2008 15:11:49 -0000 1.2
--- ld/testsuite/ld-scripts/rgn-over8-ok.d 18 Nov 2009 16:35:26 -0000
***************
*** 2,7 ****
--- 2,10 ----
# source: rgn-over8.s
# ld: -T rgn-over8.t
# objdump: -w -h
+ # xfail: rx-*-*
+ # FAILS on the RX because the linker has to set LMA == VMA for the
+ # Renesas loader.
.*: file format .*