* [BUG] perf test 21("Test object code reading") failure on ARM64
@ 2015-12-17 11:41 xiakaixu
2015-12-17 12:09 ` Jiri Olsa
0 siblings, 1 reply; 9+ messages in thread
From: xiakaixu @ 2015-12-17 11:41 UTC (permalink / raw)
To: adrian.hunter, Arnaldo Carvalho de Melo, Ingo Molnar, Jiri Olsa,
masami.hiramatsu.pt
Cc: linux-kernel, Wangnan (F)
Hi,
Perf test "Test object code reading" failed on ARM64 board and the test log is here.
######################################################################################
# perf test -v 21
...
File is: /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
On file address is: 0x70c3c
Objdump command is: objdump -z -d --start-address=0x70c3c --stop-address=0x70cbc /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
objdump read too few bytes
Bytes read differ from those read by objdump
buf1 (dso):
0x16 0x58 0x41 0xb9 0xf6 0x0e 0x00 0x35 0x00 0x00 0x00 0x90 0xbf 0x27 0x00 0xf9
0x00 0x40 0x22 0x91 0xa0 0x23 0x00 0xf9 0xb4 0x06 0x00 0xf0 0x55 0xd0 0x3b 0xd5
0x80 0xe2 0x14 0x91 0xb5 0xc2 0x1b 0xd1 0x00 0x60 0x00 0x91 0x01 0x04 0x40 0xf9
0x3f 0x00 0x15 0xeb 0xc0 0x01 0x00 0x54 0xbf 0x3f 0x00 0xb9 0x21 0x00 0x80 0x52
0x02 0xfc 0x5f 0x88 0x5f 0x00 0x1f 0x6b 0x61 0x00 0x00 0x54 0x01 0x7c 0x03 0x88
0x83 0xff 0xff 0x35 0x40 0x00 0x00 0x54 0xa2 0x3f 0x00 0xb9 0xa0 0x3f 0x40 0xb9
0xc0 0x0f 0x00 0x35 0x80 0xe2 0x14 0x91 0x15 0x10 0x00 0xf9 0x80 0xe2 0x14 0x91
0x62 0x02 0x40 0xb9 0x44 0x00 0x11 0x12 0x01 0x1c 0x40 0xb9 0x13 0x08 0x00 0xf9
buf2 (objdump):
0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
test child finished with -1
---- end ----
Test object code reading: FAILED!
######################################################################################
The result of objdump on ARM64 like this:
...
70c40: 35000ef6 cbnz w22, 70e1c <_IO_link_in+0x208>
70c44: 90000000 adrp x0, 70000 <_IO_file_fopen+0x164>
70c48: f90027bf str xzr, [x29,#72]
70c4c: 91224000 add x0, x0, #0x890
70c50: f90023a0 str x0, [x29,#64]
70c54: f00006b4 adrp x20, 147000 <__abort_msg+0x588>
...
So the following patch is needed.
diff --git a/tools/perf/tests/code-reading.c b/tools/perf/tests/code-reading.c
index a767a64..1b55fa0 100644
--- a/tools/perf/tests/code-reading.c
+++ b/tools/perf/tests/code-reading.c
@@ -61,9 +61,6 @@ static size_t read_objdump_line(const char *line, size_t line_len, void *buf,
if (i >= line_len || !isxdigit(line[i]))
break;
c2 = line[i++];
- /* Followed by a space */
- if (i < line_len && line[i] && !isspace(line[i]))
- break;
/* Store byte */
*(unsigned char *)buf = (hex(c1) << 4) | hex(c2);
buf += 1;
After applying this patch, the test still failed.
######################################################################################
...
Objdump command is: objdump -z -d --start-address=0x7c4c4 --stop-address=0x7c544 /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
Bytes read differ from those read by objdump
buf1 (dso):
0x00 0x00 0x80 0xd2 0xd5 0xff 0xff 0x17 0xe0 0x03 0x19 0xaa 0xd3 0xff 0xff 0x17
0xe1 0x03 0x14 0xaa 0xa2 0x63 0x02 0x91 0xe0 0x03 0x13 0xaa 0x66 0xfe 0xff 0x97
0xfc 0x03 0x00 0xaa 0xa0 0x4f 0x40 0xf9 0xe2 0x03 0x1c 0xaa 0xe1 0x03 0x00 0xaa
0x08 0x00 0x67 0x9e 0x61 0x02 0x01 0x8b 0xe0 0x03 0x13 0xaa 0x60 0x01 0x00 0x94
0xe0 0xf9 0xff 0x35 0x95 0x07 0x00 0xd1 0x1b 0x00 0x80 0xd2 0x01 0x01 0x66 0x9e
0x60 0x02 0x15 0x8b 0x17 0x00 0x1c 0xcb 0xf8 0x03 0x1b 0xaa 0x0a 0x00 0x67 0x9e
0x20 0x00 0x80 0xd2 0x00 0x00 0x1c 0xcb 0x81 0x02 0x01 0xcb 0x09 0x00 0x67 0x9e
0x2b 0x00 0x67 0x9e 0x16 0x03 0x14 0x8b 0x20 0x03 0x1a 0x8b 0x01 0x00 0x80 0x52
buf2 (objdump):
0xd2 0x80 0x00 0x00 0x17 0xff 0xff 0xd5 0xaa 0x19 0x03 0xe0 0x17 0xff 0xff 0xd3
0xaa 0x14 0x03 0xe1 0x91 0x02 0x63 0xa2 0xaa 0x13 0x03 0xe0 0x97 0xff 0xfe 0x66
0xaa 0x00 0x03 0xfc 0xf9 0x40 0x4f 0xa0 0xaa 0x1c 0x03 0xe2 0xaa 0x00 0x03 0xe1
0x9e 0x67 0x00 0x08 0x8b 0x01 0x02 0x61 0xaa 0x13 0x03 0xe0 0x94 0x00 0x01 0x60
0x35 0xff 0xf9 0xe0 0xd1 0x00 0x07 0x95 0xd2 0x80 0x00 0x1b 0x9e 0x66 0x01 0x01
0x8b 0x15 0x02 0x60 0xcb 0x1c 0x00 0x17 0xaa 0x1b 0x03 0xf8 0x9e 0x67 0x00 0x0a
0xd2 0x80 0x00 0x20 0xcb 0x1c 0x00 0x00 0xcb 0x01 0x02 0x81 0x9e 0x67 0x00 0x09
0x9e 0x67 0x00 0x2b 0x8b 0x14 0x03 0x16 0x8b 0x1a 0x03 0x20 0x52 0x80 0x00 0x01
test child finished with -1
---- end ----
Test object code reading: FAILED!
######################################################################################
Seems the dso file format is different between x86 and ARM64.
Maybe this test case only works fine on x86.
--
Regards
Kaixu Xia
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [BUG] perf test 21("Test object code reading") failure on ARM64
2015-12-17 11:41 [BUG] perf test 21("Test object code reading") failure on ARM64 xiakaixu
@ 2015-12-17 12:09 ` Jiri Olsa
2015-12-17 12:24 ` Jan Stancek
0 siblings, 1 reply; 9+ messages in thread
From: Jiri Olsa @ 2015-12-17 12:09 UTC (permalink / raw)
To: xiakaixu
Cc: adrian.hunter, Arnaldo Carvalho de Melo, Ingo Molnar,
masami.hiramatsu.pt, linux-kernel, Wangnan (F),
Jan Stancek
adding Jan Stancek to the loop
jirka
On Thu, Dec 17, 2015 at 07:41:43PM +0800, xiakaixu wrote:
> Hi,
>
> Perf test "Test object code reading" failed on ARM64 board and the test log is here.
>
> ######################################################################################
> # perf test -v 21
> ...
> File is: /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
> On file address is: 0x70c3c
> Objdump command is: objdump -z -d --start-address=0x70c3c --stop-address=0x70cbc /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
> objdump read too few bytes
> Bytes read differ from those read by objdump
> buf1 (dso):
> 0x16 0x58 0x41 0xb9 0xf6 0x0e 0x00 0x35 0x00 0x00 0x00 0x90 0xbf 0x27 0x00 0xf9
> 0x00 0x40 0x22 0x91 0xa0 0x23 0x00 0xf9 0xb4 0x06 0x00 0xf0 0x55 0xd0 0x3b 0xd5
> 0x80 0xe2 0x14 0x91 0xb5 0xc2 0x1b 0xd1 0x00 0x60 0x00 0x91 0x01 0x04 0x40 0xf9
> 0x3f 0x00 0x15 0xeb 0xc0 0x01 0x00 0x54 0xbf 0x3f 0x00 0xb9 0x21 0x00 0x80 0x52
> 0x02 0xfc 0x5f 0x88 0x5f 0x00 0x1f 0x6b 0x61 0x00 0x00 0x54 0x01 0x7c 0x03 0x88
> 0x83 0xff 0xff 0x35 0x40 0x00 0x00 0x54 0xa2 0x3f 0x00 0xb9 0xa0 0x3f 0x40 0xb9
> 0xc0 0x0f 0x00 0x35 0x80 0xe2 0x14 0x91 0x15 0x10 0x00 0xf9 0x80 0xe2 0x14 0x91
> 0x62 0x02 0x40 0xb9 0x44 0x00 0x11 0x12 0x01 0x1c 0x40 0xb9 0x13 0x08 0x00 0xf9
>
> buf2 (objdump):
> 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
>
> test child finished with -1
> ---- end ----
> Test object code reading: FAILED!
> ######################################################################################
>
> The result of objdump on ARM64 like this:
> ...
> 70c40: 35000ef6 cbnz w22, 70e1c <_IO_link_in+0x208>
> 70c44: 90000000 adrp x0, 70000 <_IO_file_fopen+0x164>
> 70c48: f90027bf str xzr, [x29,#72]
> 70c4c: 91224000 add x0, x0, #0x890
> 70c50: f90023a0 str x0, [x29,#64]
> 70c54: f00006b4 adrp x20, 147000 <__abort_msg+0x588>
> ...
>
> So the following patch is needed.
> diff --git a/tools/perf/tests/code-reading.c b/tools/perf/tests/code-reading.c
> index a767a64..1b55fa0 100644
> --- a/tools/perf/tests/code-reading.c
> +++ b/tools/perf/tests/code-reading.c
> @@ -61,9 +61,6 @@ static size_t read_objdump_line(const char *line, size_t line_len, void *buf,
> if (i >= line_len || !isxdigit(line[i]))
> break;
> c2 = line[i++];
> - /* Followed by a space */
> - if (i < line_len && line[i] && !isspace(line[i]))
> - break;
> /* Store byte */
> *(unsigned char *)buf = (hex(c1) << 4) | hex(c2);
> buf += 1;
>
> After applying this patch, the test still failed.
> ######################################################################################
> ...
> Objdump command is: objdump -z -d --start-address=0x7c4c4 --stop-address=0x7c544 /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
> Bytes read differ from those read by objdump
> buf1 (dso):
> 0x00 0x00 0x80 0xd2 0xd5 0xff 0xff 0x17 0xe0 0x03 0x19 0xaa 0xd3 0xff 0xff 0x17
> 0xe1 0x03 0x14 0xaa 0xa2 0x63 0x02 0x91 0xe0 0x03 0x13 0xaa 0x66 0xfe 0xff 0x97
> 0xfc 0x03 0x00 0xaa 0xa0 0x4f 0x40 0xf9 0xe2 0x03 0x1c 0xaa 0xe1 0x03 0x00 0xaa
> 0x08 0x00 0x67 0x9e 0x61 0x02 0x01 0x8b 0xe0 0x03 0x13 0xaa 0x60 0x01 0x00 0x94
> 0xe0 0xf9 0xff 0x35 0x95 0x07 0x00 0xd1 0x1b 0x00 0x80 0xd2 0x01 0x01 0x66 0x9e
> 0x60 0x02 0x15 0x8b 0x17 0x00 0x1c 0xcb 0xf8 0x03 0x1b 0xaa 0x0a 0x00 0x67 0x9e
> 0x20 0x00 0x80 0xd2 0x00 0x00 0x1c 0xcb 0x81 0x02 0x01 0xcb 0x09 0x00 0x67 0x9e
> 0x2b 0x00 0x67 0x9e 0x16 0x03 0x14 0x8b 0x20 0x03 0x1a 0x8b 0x01 0x00 0x80 0x52
>
> buf2 (objdump):
> 0xd2 0x80 0x00 0x00 0x17 0xff 0xff 0xd5 0xaa 0x19 0x03 0xe0 0x17 0xff 0xff 0xd3
> 0xaa 0x14 0x03 0xe1 0x91 0x02 0x63 0xa2 0xaa 0x13 0x03 0xe0 0x97 0xff 0xfe 0x66
> 0xaa 0x00 0x03 0xfc 0xf9 0x40 0x4f 0xa0 0xaa 0x1c 0x03 0xe2 0xaa 0x00 0x03 0xe1
> 0x9e 0x67 0x00 0x08 0x8b 0x01 0x02 0x61 0xaa 0x13 0x03 0xe0 0x94 0x00 0x01 0x60
> 0x35 0xff 0xf9 0xe0 0xd1 0x00 0x07 0x95 0xd2 0x80 0x00 0x1b 0x9e 0x66 0x01 0x01
> 0x8b 0x15 0x02 0x60 0xcb 0x1c 0x00 0x17 0xaa 0x1b 0x03 0xf8 0x9e 0x67 0x00 0x0a
> 0xd2 0x80 0x00 0x20 0xcb 0x1c 0x00 0x00 0xcb 0x01 0x02 0x81 0x9e 0x67 0x00 0x09
> 0x9e 0x67 0x00 0x2b 0x8b 0x14 0x03 0x16 0x8b 0x1a 0x03 0x20 0x52 0x80 0x00 0x01
>
> test child finished with -1
> ---- end ----
> Test object code reading: FAILED!
> ######################################################################################
>
> Seems the dso file format is different between x86 and ARM64.
> Maybe this test case only works fine on x86.
> --
> Regards
> Kaixu Xia
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [BUG] perf test 21("Test object code reading") failure on ARM64
2015-12-17 12:09 ` Jiri Olsa
@ 2015-12-17 12:24 ` Jan Stancek
2015-12-19 3:04 ` xiakaixu
0 siblings, 1 reply; 9+ messages in thread
From: Jan Stancek @ 2015-12-17 12:24 UTC (permalink / raw)
To: Jiri Olsa, xiakaixu
Cc: adrian hunter, Arnaldo Carvalho de Melo, Ingo Molnar,
masami hiramatsu pt, linux-kernel, Wangnan (F)
----- Original Message -----
> From: "Jiri Olsa" <jolsa@redhat.com>
> To: "xiakaixu" <xiakaixu@huawei.com>
> Cc: "adrian hunter" <adrian.hunter@intel.com>, "Arnaldo Carvalho de Melo" <acme@kernel.org>, "Ingo Molnar"
> <mingo@kernel.org>, "masami hiramatsu pt" <masami.hiramatsu.pt@hitachi.com>, linux-kernel@vger.kernel.org, "Wangnan
> (F)" <wangnan0@huawei.com>, "Jan Stancek" <jstancek@redhat.com>
> Sent: Thursday, 17 December, 2015 1:09:26 PM
> Subject: Re: [BUG] perf test 21("Test object code reading") failure on ARM64
>
> adding Jan Stancek to the loop
>
> jirka
>
> On Thu, Dec 17, 2015 at 07:41:43PM +0800, xiakaixu wrote:
> > Hi,
> >
> > Perf test "Test object code reading" failed on ARM64 board and the test log
> > is here.
> >
> > ######################################################################################
> > # perf test -v 21
> > ...
> > File is: /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
> > On file address is: 0x70c3c
> > Objdump command is: objdump -z -d --start-address=0x70c3c
> > --stop-address=0x70cbc /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
> > objdump read too few bytes
> > Bytes read differ from those read by objdump
> > buf1 (dso):
> > 0x16 0x58 0x41 0xb9 0xf6 0x0e 0x00 0x35 0x00 0x00 0x00 0x90 0xbf 0x27
> > 0x00 0xf9
> > 0x00 0x40 0x22 0x91 0xa0 0x23 0x00 0xf9 0xb4 0x06 0x00 0xf0 0x55 0xd0
> > 0x3b 0xd5
> > 0x80 0xe2 0x14 0x91 0xb5 0xc2 0x1b 0xd1 0x00 0x60 0x00 0x91 0x01 0x04
> > 0x40 0xf9
> > 0x3f 0x00 0x15 0xeb 0xc0 0x01 0x00 0x54 0xbf 0x3f 0x00 0xb9 0x21 0x00
> > 0x80 0x52
> > 0x02 0xfc 0x5f 0x88 0x5f 0x00 0x1f 0x6b 0x61 0x00 0x00 0x54 0x01 0x7c
> > 0x03 0x88
> > 0x83 0xff 0xff 0x35 0x40 0x00 0x00 0x54 0xa2 0x3f 0x00 0xb9 0xa0 0x3f
> > 0x40 0xb9
> > 0xc0 0x0f 0x00 0x35 0x80 0xe2 0x14 0x91 0x15 0x10 0x00 0xf9 0x80 0xe2
> > 0x14 0x91
> > 0x62 0x02 0x40 0xb9 0x44 0x00 0x11 0x12 0x01 0x1c 0x40 0xb9 0x13 0x08
> > 0x00 0xf9
> >
> > buf2 (objdump):
> > 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> > 0x00 0x00
> > 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> > 0x00 0x00
> > 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> > 0x00 0x00
> > 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> > 0x00 0x00
> > 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> > 0x00 0x00
> > 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> > 0x00 0x00
> > 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> > 0x00 0x00
> > 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00
> > 0x00 0x00
> >
> > test child finished with -1
> > ---- end ----
> > Test object code reading: FAILED!
> > ######################################################################################
> >
> > The result of objdump on ARM64 like this:
> > ...
> > 70c40: 35000ef6 cbnz w22, 70e1c <_IO_link_in+0x208>
> > 70c44: 90000000 adrp x0, 70000 <_IO_file_fopen+0x164>
> > 70c48: f90027bf str xzr, [x29,#72]
> > 70c4c: 91224000 add x0, x0, #0x890
> > 70c50: f90023a0 str x0, [x29,#64]
> > 70c54: f00006b4 adrp x20, 147000 <__abort_msg+0x588>
> > ...
Hi,
What is your objdump version?
> >
> > So the following patch is needed.
> > diff --git a/tools/perf/tests/code-reading.c
> > b/tools/perf/tests/code-reading.c
> > index a767a64..1b55fa0 100644
> > --- a/tools/perf/tests/code-reading.c
> > +++ b/tools/perf/tests/code-reading.c
> > @@ -61,9 +61,6 @@ static size_t read_objdump_line(const char *line, size_t
> > line_len, void *buf,
> > if (i >= line_len || !isxdigit(line[i]))
> > break;
> > c2 = line[i++];
> > - /* Followed by a space */
> > - if (i < line_len && line[i] && !isspace(line[i]))
> > - break;
> > /* Store byte */
> > *(unsigned char *)buf = (hex(c1) << 4) | hex(c2);
> > buf += 1;
> >
> > After applying this patch, the test still failed.
> > ######################################################################################
> > ...
> > Objdump command is: objdump -z -d --start-address=0x7c4c4
> > --stop-address=0x7c544 /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
> > Bytes read differ from those read by objdump
> > buf1 (dso):
> > 0x00 0x00 0x80 0xd2 0xd5 0xff 0xff 0x17 0xe0 0x03 0x19 0xaa 0xd3 0xff
> > 0xff 0x17
> > 0xe1 0x03 0x14 0xaa 0xa2 0x63 0x02 0x91 0xe0 0x03 0x13 0xaa 0x66 0xfe
> > 0xff 0x97
> > 0xfc 0x03 0x00 0xaa 0xa0 0x4f 0x40 0xf9 0xe2 0x03 0x1c 0xaa 0xe1 0x03
> > 0x00 0xaa
> > 0x08 0x00 0x67 0x9e 0x61 0x02 0x01 0x8b 0xe0 0x03 0x13 0xaa 0x60 0x01
> > 0x00 0x94
> > 0xe0 0xf9 0xff 0x35 0x95 0x07 0x00 0xd1 0x1b 0x00 0x80 0xd2 0x01 0x01
> > 0x66 0x9e
> > 0x60 0x02 0x15 0x8b 0x17 0x00 0x1c 0xcb 0xf8 0x03 0x1b 0xaa 0x0a 0x00
> > 0x67 0x9e
> > 0x20 0x00 0x80 0xd2 0x00 0x00 0x1c 0xcb 0x81 0x02 0x01 0xcb 0x09 0x00
> > 0x67 0x9e
> > 0x2b 0x00 0x67 0x9e 0x16 0x03 0x14 0x8b 0x20 0x03 0x1a 0x8b 0x01 0x00
> > 0x80 0x52
> >
> > buf2 (objdump):
> > 0xd2 0x80 0x00 0x00 0x17 0xff 0xff 0xd5 0xaa 0x19 0x03 0xe0 0x17 0xff
> > 0xff 0xd3
> > 0xaa 0x14 0x03 0xe1 0x91 0x02 0x63 0xa2 0xaa 0x13 0x03 0xe0 0x97 0xff
> > 0xfe 0x66
> > 0xaa 0x00 0x03 0xfc 0xf9 0x40 0x4f 0xa0 0xaa 0x1c 0x03 0xe2 0xaa 0x00
> > 0x03 0xe1
> > 0x9e 0x67 0x00 0x08 0x8b 0x01 0x02 0x61 0xaa 0x13 0x03 0xe0 0x94 0x00
> > 0x01 0x60
> > 0x35 0xff 0xf9 0xe0 0xd1 0x00 0x07 0x95 0xd2 0x80 0x00 0x1b 0x9e 0x66
> > 0x01 0x01
> > 0x8b 0x15 0x02 0x60 0xcb 0x1c 0x00 0x17 0xaa 0x1b 0x03 0xf8 0x9e 0x67
> > 0x00 0x0a
> > 0xd2 0x80 0x00 0x20 0xcb 0x1c 0x00 0x00 0xcb 0x01 0x02 0x81 0x9e 0x67
> > 0x00 0x09
> > 0x9e 0x67 0x00 0x2b 0x8b 0x14 0x03 0x16 0x8b 0x1a 0x03 0x20 0x52 0x80
> > 0x00 0x01
The data appears to match, but the endian is different.
Regards,
Jan
> >
> > test child finished with -1
> > ---- end ----
> > Test object code reading: FAILED!
> > ######################################################################################
> >
> > Seems the dso file format is different between x86 and ARM64.
> > Maybe this test case only works fine on x86.
> > --
> > Regards
> > Kaixu Xia
> >
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [BUG] perf test 21("Test object code reading") failure on ARM64
2015-12-17 12:24 ` Jan Stancek
@ 2015-12-19 3:04 ` xiakaixu
2015-12-20 0:25 ` Jan Stancek
0 siblings, 1 reply; 9+ messages in thread
From: xiakaixu @ 2015-12-19 3:04 UTC (permalink / raw)
To: Jan Stancek
Cc: Jiri Olsa, adrian hunter, Arnaldo Carvalho de Melo, Ingo Molnar,
masami hiramatsu pt, linux-kernel, Wangnan (F)
>>> ...
>
> Hi,
>
> What is your objdump version?
Hi,
Sorry for the late reply.
# objdump --version
GNU objdump (GNU Binutils) 2.25.
I am sure that the system is Little endian.
>
>>>
>>> So the following patch is needed.
>>> diff --git a/tools/perf/tests/code-reading.c
>>> b/tools/perf/tests/code-reading.c
>>> index a767a64..1b55fa0 100644
>>> --- a/tools/perf/tests/code-reading.c
>>> +++ b/tools/perf/tests/code-reading.c
>>> @@ -61,9 +61,6 @@ static size_t read_objdump_line(const char *line, size_t
>>> line_len, void *buf,
>>> if (i >= line_len || !isxdigit(line[i]))
>>> break;
>>> c2 = line[i++];
>>> - /* Followed by a space */
>>> - if (i < line_len && line[i] && !isspace(line[i]))
>>> - break;
>>> /* Store byte */
>>> *(unsigned char *)buf = (hex(c1) << 4) | hex(c2);
>>> buf += 1;
>>>
>>> After applying this patch, the test still failed.
>>> ######################################################################################
>>> ...
>>> Objdump command is: objdump -z -d --start-address=0x7c4c4
>>> --stop-address=0x7c544 /tmp/oxygen_root-root/lib64/libc-2.19-2014.08.so
>>> Bytes read differ from those read by objdump
>>> buf1 (dso):
>>> 0x00 0x00 0x80 0xd2 0xd5 0xff 0xff 0x17 0xe0 0x03 0x19 0xaa 0xd3 0xff
>>> 0xff 0x17
>>> 0xe1 0x03 0x14 0xaa 0xa2 0x63 0x02 0x91 0xe0 0x03 0x13 0xaa 0x66 0xfe
>>> 0xff 0x97
>>> 0xfc 0x03 0x00 0xaa 0xa0 0x4f 0x40 0xf9 0xe2 0x03 0x1c 0xaa 0xe1 0x03
>>> 0x00 0xaa
>>> 0x08 0x00 0x67 0x9e 0x61 0x02 0x01 0x8b 0xe0 0x03 0x13 0xaa 0x60 0x01
>>> 0x00 0x94
>>> 0xe0 0xf9 0xff 0x35 0x95 0x07 0x00 0xd1 0x1b 0x00 0x80 0xd2 0x01 0x01
>>> 0x66 0x9e
>>> 0x60 0x02 0x15 0x8b 0x17 0x00 0x1c 0xcb 0xf8 0x03 0x1b 0xaa 0x0a 0x00
>>> 0x67 0x9e
>>> 0x20 0x00 0x80 0xd2 0x00 0x00 0x1c 0xcb 0x81 0x02 0x01 0xcb 0x09 0x00
>>> 0x67 0x9e
>>> 0x2b 0x00 0x67 0x9e 0x16 0x03 0x14 0x8b 0x20 0x03 0x1a 0x8b 0x01 0x00
>>> 0x80 0x52
>>>
>>> buf2 (objdump):
>>> 0xd2 0x80 0x00 0x00 0x17 0xff 0xff 0xd5 0xaa 0x19 0x03 0xe0 0x17 0xff
>>> 0xff 0xd3
>>> 0xaa 0x14 0x03 0xe1 0x91 0x02 0x63 0xa2 0xaa 0x13 0x03 0xe0 0x97 0xff
>>> 0xfe 0x66
>>> 0xaa 0x00 0x03 0xfc 0xf9 0x40 0x4f 0xa0 0xaa 0x1c 0x03 0xe2 0xaa 0x00
>>> 0x03 0xe1
>>> 0x9e 0x67 0x00 0x08 0x8b 0x01 0x02 0x61 0xaa 0x13 0x03 0xe0 0x94 0x00
>>> 0x01 0x60
>>> 0x35 0xff 0xf9 0xe0 0xd1 0x00 0x07 0x95 0xd2 0x80 0x00 0x1b 0x9e 0x66
>>> 0x01 0x01
>>> 0x8b 0x15 0x02 0x60 0xcb 0x1c 0x00 0x17 0xaa 0x1b 0x03 0xf8 0x9e 0x67
>>> 0x00 0x0a
>>> 0xd2 0x80 0x00 0x20 0xcb 0x1c 0x00 0x00 0xcb 0x01 0x02 0x81 0x9e 0x67
>>> 0x00 0x09
>>> 0x9e 0x67 0x00 0x2b 0x8b 0x14 0x03 0x16 0x8b 0x1a 0x03 0x20 0x52 0x80
>>> 0x00 0x01
>
> The data appears to match, but the endian is different.
>
> Regards,
> Jan
>
>>>
>>> test child finished with -1
>>> ---- end ----
>>> Test object code reading: FAILED!
>>> ######################################################################################
>>>
>>> Seems the dso file format is different between x86 and ARM64.
>>> Maybe this test case only works fine on x86.
>>> --
>>> Regards
>>> Kaixu Xia
>>>
>>
>
> .
>
--
Regards
Kaixu Xia
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [BUG] perf test 21("Test object code reading") failure on ARM64
2015-12-19 3:04 ` xiakaixu
@ 2015-12-20 0:25 ` Jan Stancek
2015-12-21 2:18 ` xiakaixu
0 siblings, 1 reply; 9+ messages in thread
From: Jan Stancek @ 2015-12-20 0:25 UTC (permalink / raw)
To: xiakaixu
Cc: Jiri Olsa, adrian hunter, Arnaldo Carvalho de Melo, Ingo Molnar,
masami hiramatsu pt, linux-kernel, Wangnan (F)
[-- Attachment #1: Type: text/plain, Size: 442 bytes --]
On Sat, Dec 19, 2015 at 11:04:21AM +0800, xiakaixu wrote:
>
> >>> ...
> >
> > Hi,
> >
> > What is your objdump version?
>
> Hi,
>
> Sorry for the late reply.
>
> # objdump --version
> GNU objdump (GNU Binutils) 2.25.
>
> I am sure that the system is Little endian.
> >
I have attached a patch if you care to try it with your setup.
If it still fails, output from -v and last objdump command output
would be helpful.
Regards,
Jan
[-- Attachment #2: code_reading_in_chunks.patch --]
[-- Type: text/plain, Size: 3120 bytes --]
diff --git a/tools/perf/tests/code-reading.c b/tools/perf/tests/code-reading.c
index a767a6400c5c..faf726865fac 100644
--- a/tools/perf/tests/code-reading.c
+++ b/tools/perf/tests/code-reading.c
@@ -33,44 +33,83 @@ static unsigned int hex(char c)
return c - 'A' + 10;
}
-static size_t read_objdump_line(const char *line, size_t line_len, void *buf,
- size_t len)
+static size_t read_objdump_chunk(const char **line, unsigned char **buf,
+ size_t *buf_len)
+{
+ size_t bytes_read = 0;
+ unsigned char *chunk_start = *buf;
+
+ /* Read bytes */
+ while (*buf_len > 0) {
+ char c1, c2;
+
+ /* Get 2 hex digits */
+ c1 = *(*line)++;
+ if (!isxdigit(c1))
+ break;
+ c2 = *(*line)++;
+ if (!isxdigit(c2))
+ break;
+
+ /* Store byte and advance buf */
+ **buf = (hex(c1) << 4) | hex(c2);
+ (*buf)++;
+ (*buf_len)--;
+ bytes_read++;
+
+ /* End of chunk? */
+ if (isspace(**line))
+ break;
+ }
+
+ /*
+ * objdump will display raw insn as LE if code endian
+ * is LE and bytes_per_chunk > 1. In that case reverse
+ * the chunk we just read.
+ */
+ if (bytes_read > 1 && !bigendian()) {
+ unsigned char *chunk_end = chunk_start + bytes_read - 1;
+ unsigned char tmp;
+
+ while (chunk_start < chunk_end) {
+ tmp = *chunk_start;
+ *chunk_start = *chunk_end;
+ *chunk_end = tmp;
+ chunk_start++;
+ chunk_end--;
+ }
+ }
+
+ return bytes_read;
+}
+
+static size_t read_objdump_line(const char *line, unsigned char *buf,
+ size_t buf_len)
{
const char *p;
- size_t i, j = 0;
+ size_t ret, bytes_read = 0;
/* Skip to a colon */
p = strchr(line, ':');
if (!p)
return 0;
- i = p + 1 - line;
+ p++;
- /* Read bytes */
- while (j < len) {
- char c1, c2;
-
- /* Skip spaces */
- for (; i < line_len; i++) {
- if (!isspace(line[i]))
- break;
- }
- /* Get 2 hex digits */
- if (i >= line_len || !isxdigit(line[i]))
- break;
- c1 = line[i++];
- if (i >= line_len || !isxdigit(line[i]))
- break;
- c2 = line[i++];
- /* Followed by a space */
- if (i < line_len && line[i] && !isspace(line[i]))
+ /* Skip initial spaces */
+ while (*p) {
+ if (!isspace(*p))
break;
- /* Store byte */
- *(unsigned char *)buf = (hex(c1) << 4) | hex(c2);
- buf += 1;
- j++;
+ p++;
}
+
+ do {
+ ret = read_objdump_chunk(&p, &buf, &buf_len);
+ bytes_read += ret;
+ p++;
+ } while (ret > 0);
+
/* return number of successfully read bytes */
- return j;
+ return bytes_read;
}
static int read_objdump_output(FILE *f, void *buf, size_t *len, u64 start_addr)
@@ -95,7 +134,7 @@ static int read_objdump_output(FILE *f, void *buf, size_t *len, u64 start_addr)
}
/* read objdump data into temporary buffer */
- read_bytes = read_objdump_line(line, ret, tmp, sizeof(tmp));
+ read_bytes = read_objdump_line(line, tmp, sizeof(tmp));
if (!read_bytes)
continue;
@@ -152,7 +191,7 @@ static int read_via_objdump(const char *filename, u64 addr, void *buf,
ret = read_objdump_output(f, buf, &len, addr);
if (len) {
- pr_debug("objdump read too few bytes\n");
+ pr_debug("objdump read too few bytes: %lu\n", len);
if (!ret)
ret = len;
}
^ permalink raw reply related [flat|nested] 9+ messages in thread
* Re: [BUG] perf test 21("Test object code reading") failure on ARM64
2015-12-20 0:25 ` Jan Stancek
@ 2015-12-21 2:18 ` xiakaixu
2016-01-07 16:27 ` Arnaldo Carvalho de Melo
0 siblings, 1 reply; 9+ messages in thread
From: xiakaixu @ 2015-12-21 2:18 UTC (permalink / raw)
To: Jan Stancek
Cc: Jiri Olsa, adrian hunter, Arnaldo Carvalho de Melo, Ingo Molnar,
masami hiramatsu pt, linux-kernel, Wangnan (F)
于 2015/12/20 8:25, Jan Stancek 写道:
> On Sat, Dec 19, 2015 at 11:04:21AM +0800, xiakaixu wrote:
>>
>>>>> ...
>>>
>>> Hi,
>>>
>>> What is your objdump version?
>>
>> Hi,
>>
>> Sorry for the late reply.
>>
>> # objdump --version
>> GNU objdump (GNU Binutils) 2.25.
>>
>> I am sure that the system is Little endian.
>>>
>
> I have attached a patch if you care to try it with your setup.
> If it still fails, output from -v and last objdump command output
> would be helpful.
Hi,
After applying this patch, the perf test case passed.
# perf test 21
21: Test object code reading : (no vmlinux) Ok
Thanks!
>
> Regards,
> Jan
>
--
Regards
Kaixu Xia
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [BUG] perf test 21("Test object code reading") failure on ARM64
2015-12-21 2:18 ` xiakaixu
@ 2016-01-07 16:27 ` Arnaldo Carvalho de Melo
2016-01-08 8:03 ` Jan Stancek
0 siblings, 1 reply; 9+ messages in thread
From: Arnaldo Carvalho de Melo @ 2016-01-07 16:27 UTC (permalink / raw)
To: xiakaixu
Cc: Jan Stancek, Jiri Olsa, adrian hunter, Ingo Molnar,
masami hiramatsu pt, linux-kernel, Wangnan (F)
Em Mon, Dec 21, 2015 at 10:18:20AM +0800, xiakaixu escreveu:
> 于 2015/12/20 8:25, Jan Stancek 写道:
> > On Sat, Dec 19, 2015 at 11:04:21AM +0800, xiakaixu wrote:
> >>
> >>>>> ...
> >>>
> >>> Hi,
> >>>
> >>> What is your objdump version?
> >>
> >> Hi,
> >>
> >> Sorry for the late reply.
> >>
> >> # objdump --version
> >> GNU objdump (GNU Binutils) 2.25.
> >>
> >> I am sure that the system is Little endian.
> >>>
> >
> > I have attached a patch if you care to try it with your setup.
> > If it still fails, output from -v and last objdump command output
> > would be helpful.
>
> Hi,
>
> After applying this patch, the perf test case passed.
>
> # perf test 21
> 21: Test object code reading : (no vmlinux) Ok
So, assuming this doesn't break other running the test in other arches,
can Jan please resend this with a changelog and a Tested-by: Xia?
Was that done already? /me still digging thru my queue after last month
vacations...
- Arnaldo
> Thanks!
> >
> > Regards,
> > Jan
> >
>
>
> --
> Regards
> Kaixu Xia
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [BUG] perf test 21("Test object code reading") failure on ARM64
2016-01-07 16:27 ` Arnaldo Carvalho de Melo
@ 2016-01-08 8:03 ` Jan Stancek
2016-01-08 12:52 ` Arnaldo Carvalho de Melo
0 siblings, 1 reply; 9+ messages in thread
From: Jan Stancek @ 2016-01-08 8:03 UTC (permalink / raw)
To: Arnaldo Carvalho de Melo
Cc: xiakaixu, Jiri Olsa, adrian hunter, Ingo Molnar,
masami hiramatsu pt, linux-kernel, Wangnan (F)
----- Original Message -----
> From: "Arnaldo Carvalho de Melo" <acme@kernel.org>
> To: "xiakaixu" <xiakaixu@huawei.com>
> Cc: "Jan Stancek" <jstancek@redhat.com>, "Jiri Olsa" <jolsa@redhat.com>, "adrian hunter" <adrian.hunter@intel.com>,
> "Ingo Molnar" <mingo@kernel.org>, "masami hiramatsu pt" <masami.hiramatsu.pt@hitachi.com>,
> linux-kernel@vger.kernel.org, "Wangnan (F)" <wangnan0@huawei.com>
> Sent: Thursday, 7 January, 2016 5:27:21 PM
> Subject: Re: [BUG] perf test 21("Test object code reading") failure on ARM64
>
> Em Mon, Dec 21, 2015 at 10:18:20AM +0800, xiakaixu escreveu:
> > 于 2015/12/20 8:25, Jan Stancek 写道:
> > > On Sat, Dec 19, 2015 at 11:04:21AM +0800, xiakaixu wrote:
> > >>
> > >>>>> ...
> > >>>
> > >>> Hi,
> > >>>
> > >>> What is your objdump version?
> > >>
> > >> Hi,
> > >>
> > >> Sorry for the late reply.
> > >>
> > >> # objdump --version
> > >> GNU objdump (GNU Binutils) 2.25.
> > >>
> > >> I am sure that the system is Little endian.
> > >>>
> > >
> > > I have attached a patch if you care to try it with your setup.
> > > If it still fails, output from -v and last objdump command output
> > > would be helpful.
> >
> > Hi,
> >
> > After applying this patch, the perf test case passed.
> >
> > # perf test 21
> > 21: Test object code reading : (no vmlinux)
> > Ok
>
> So, assuming this doesn't break other running the test in other arches,
> can Jan please resend this with a changelog and a Tested-by: Xia?
>
> Was that done already? /me still digging thru my queue after last month
> vacations...
Hi,
not yet. I'd like to try the patch also on ppc64 BE/LE before I post it
(likely next week).
Regards,
Jan
>
> - Arnaldo
>
> > Thanks!
> > >
> > > Regards,
> > > Jan
> > >
> >
> >
> > --
> > Regards
> > Kaixu Xia
>
^ permalink raw reply [flat|nested] 9+ messages in thread
* Re: [BUG] perf test 21("Test object code reading") failure on ARM64
2016-01-08 8:03 ` Jan Stancek
@ 2016-01-08 12:52 ` Arnaldo Carvalho de Melo
0 siblings, 0 replies; 9+ messages in thread
From: Arnaldo Carvalho de Melo @ 2016-01-08 12:52 UTC (permalink / raw)
To: Jan Stancek
Cc: xiakaixu, Jiri Olsa, adrian hunter, Ingo Molnar,
masami hiramatsu pt, linux-kernel, Wangnan (F)
Em Fri, Jan 08, 2016 at 03:03:34AM -0500, Jan Stancek escreveu:
> > From: "Arnaldo Carvalho de Melo" <acme@kernel.org>
> > Em Mon, Dec 21, 2015 at 10:18:20AM +0800, xiakaixu escreveu:
> > > After applying this patch, the perf test case passed.
> > > # perf test 21
> > > 21: Test object code reading : (no vmlinux)
> > > Ok
> > So, assuming this doesn't break other running the test in other arches,
> > can Jan please resend this with a changelog and a Tested-by: Xia?
> > Was that done already? /me still digging thru my queue after last month
> > vacations...
> not yet. I'd like to try the patch also on ppc64 BE/LE before I post it
> (likely next week).
Thanks, I'll wait then,
- Arnaldo
^ permalink raw reply [flat|nested] 9+ messages in thread
end of thread, other threads:[~2016-01-08 12:52 UTC | newest]
Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-12-17 11:41 [BUG] perf test 21("Test object code reading") failure on ARM64 xiakaixu
2015-12-17 12:09 ` Jiri Olsa
2015-12-17 12:24 ` Jan Stancek
2015-12-19 3:04 ` xiakaixu
2015-12-20 0:25 ` Jan Stancek
2015-12-21 2:18 ` xiakaixu
2016-01-07 16:27 ` Arnaldo Carvalho de Melo
2016-01-08 8:03 ` Jan Stancek
2016-01-08 12:52 ` Arnaldo Carvalho de Melo
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.