This is the mail archive of the binutils@sourceware.org mailing list for the binutils project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: New test ld-elf/header fails for cris-elf (likely elsewhere)


Hans-Peter Nilsson wrote:

with ld.log:
./ld-new  -L/h/hp/binutils/cvs_latest/src/ld/testsuite/ld-elf  -T header.t -z max-page-size=0x10000 -o tmpdir/dump tmpdir/dump0.o
failed with: <./ld-new: unrecognized option '-z'
./ld-new: use the --help option for usage information>, expected: <>
./ld-new: unrecognized option '-z'
./ld-new: use the --help option for usage information
/n/asic_slask/hp/autobinutest/ld/../binutils/objdump  -hpw tmpdir/dump > tmpdir/dump.out
regexp_diff match failure
regexp "^    LOAD off    0x0*0000000 vaddr 0x0*0010000 paddr 0x0*0010000 align 2..16$"
line   "    LOAD off    0x00000000 vaddr 0x00000000 paddr 0x00000000 align 2**13"
regexp_diff match failure
regexp "^         filesz 0x0*001002[48] memsz 0x0*001002[48] flags rwx$"
line   "         filesz 0x00001002 memsz 0x00001002 flags rw-"
(more errors elided)

It seems this test should be restricted to linux and just for
one architecture (perhaps just for ix86-GNU/Linux), as the
expected output checks for certain alignment for sections and
for certain layout.  Depending on the scope of your test
(comment absent in header.d), I suggest you add a line at the
top of header.d:

#target: *-*-linux*

(Though, that doesn't cover the 64k-alignment issue.  Does this
test really pass on ix86-linux?)

Should we add that or more such target lines, or should header.d
be generalized regarding output?  (What's really to be tested?)

Primarily what's being tested is that the link doesn't crash. Perhaps it's just sufficient to check that, rather than also check the output.


Whether the right alignments occur to tickle the orginal bug depends on the program header size and the page size. As I don't think the page size is forceable from the test harness, or detectable in the assembler source, perhaps we should just check the link succeeds. Thoughts?

nathan

--
Nathan Sidwell    ::   http://www.codesourcery.com   ::         CodeSourcery
nathan@codesourcery.com    ::     http://www.planetfall.pwp.blueyonder.co.uk


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]