* [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read
@ 2012-06-20 10:09 Hui Zhu
2012-06-20 14:19 ` Yao Qi
2012-06-20 14:44 ` Yao Qi
0 siblings, 2 replies; 8+ messages in thread
From: Hui Zhu @ 2012-06-20 10:09 UTC (permalink / raw)
To: gdb-patches ml; +Cc: Pedro Alves
Hi,
This test is for the fix in
http://sourceware.org/ml/gdb-patches/2012-06/msg00504.html
I test it in amd64 and it work OK.
Please help me review it.
Best,
Hui
2012-06-20 Hui Zhu <hui_zhu@mentor.com>
* gdb.trace/tfile.exp: Add new test for backtrace commands
with tfile to test the tfile memory read.
---
testsuite/gdb.trace/tfile.exp | 52 ++++++++++++++++++++++++++++++++++++++++++
1 file changed, 52 insertions(+)
--- a/testsuite/gdb.trace/tfile.exp
+++ b/testsuite/gdb.trace/tfile.exp
@@ -117,3 +117,55 @@ Not looking at any trace frame.*" \
gdb_test \
"interpreter-exec mi \"-target-select tfile basic.tf\"" \
"\\^connected.*"
+
+#Test the backtrace commands with tfile to test the tfile memory read
+remote_file host delete bt.tf
+
+gdb_exit
+gdb_start
+
+gdb_load $binfile
+
+runto_main
+
+gdb_delete_tracepoints
+set tp1 [gdb_gettpnum write_error_trace_file];
+
+if { [istarget i?86-*-*] } {
+ gdb_trace_setactions "Set actions for collect stack" \
+ "$tp1" \
+ "collect *(unsigned char *)\$esp@128" "^$"
+} elseif {[istarget x86_64-*-* ]} {
+ gdb_trace_setactions "Set actions for collect stack" \
+ "$tp1" \
+ "collect *(unsigned char *)\$rsp@128" "^$"
+} elseif {[istarget arm*-*-* ] || [istarget mips*-*-* ]} {
+ gdb_trace_setactions "Set actions for collect stack" \
+ "$tp1" \
+ "collect *(unsigned char *)\$sp@128" "^$"
+} else {
+ unsupported "don't know howto use tracepoint collect stack in this target"
+ continue
+}
+
+gdb_test_no_output "tstart"
+gdb_test "break done_making_trace_files" ".*" ""
+gdb_test "continue" ".*" ""
+gdb_test_no_output "tstop"
+gdb_test "tsave ./bt.tf" ".*" ""
+
+gdb_exit
+gdb_start
+
+gdb_load $binfile
+
+gdb_test "target tfile bt.tf" "Created tracepoint.*" "load trace file bt.tf"
+gdb_test "tfind 0" \
+ "Found trace frame 0, tracepoint \[0-9\]+.
+\#0 write_error_trace_file ().*" \
+ "tfind 0 on trace file bt.tf"
+
+gdb_test "backtrace" \
+"\#0 write_error_trace_file ().*
+\#1 .* in main .*" \
+ "backtrace on trace file bt.tf"
^ permalink raw reply [flat|nested] 8+ messages in thread* Re: [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read
2012-06-20 10:09 [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read Hui Zhu
@ 2012-06-20 14:19 ` Yao Qi
2012-06-20 14:44 ` Yao Qi
1 sibling, 0 replies; 8+ messages in thread
From: Yao Qi @ 2012-06-20 14:19 UTC (permalink / raw)
To: gdb-patches; +Cc: Hui Zhu, Pedro Alves
On Wednesday 20 June 2012 18:08:17 Hui Zhu wrote:
> +#Test the backtrace commands with tfile to test the tfile memory read
> +remote_file host delete bt.tf
It is better to remove temp files when the test is about to finish.
> +
> +gdb_exit
> +gdb_start
> +
> +gdb_load $binfile
> +
These three can be replaced with "clean_restart".
> +runto_main
> +
> +gdb_delete_tracepoints
Is it needed? IIRC, all breakpoints are removed inside runto_main.
> +set tp1 [gdb_gettpnum write_error_trace_file];
> +
> +if { [istarget i?86-*-*] } {
> + gdb_trace_setactions "Set actions for collect stack" \
> + "$tp1" \
> + "collect *(unsigned char *)\$esp@128" "^$"
> +} elseif {[istarget x86_64-*-* ]} {
> + gdb_trace_setactions "Set actions for collect stack" \
> + "$tp1" \
> + "collect *(unsigned char *)\$rsp@128" "^$"
> +} elseif {[istarget arm*-*-* ] || [istarget mips*-*-* ]} {
> + gdb_trace_setactions "Set actions for collect stack" \
> + "$tp1" \
> + "collect *(unsigned char *)\$sp@128" "^$"
> +} else {
> + unsupported "don't know howto use tracepoint collect stack in this
> target" + continue
> +}
> +
> +gdb_test_no_output "tstart"
> +gdb_test "break done_making_trace_files" ".*" ""
> +gdb_test "continue" ".*" ""
> +gdb_test_no_output "tstop"
> +gdb_test "tsave ./bt.tf" ".*" ""
> +
> +gdb_exit
> +gdb_start
> +
> +gdb_load $binfile
clean_restart
> +
> +gdb_test "target tfile bt.tf" "Created tracepoint.*" "load trace file
> bt.tf" +gdb_test "tfind 0" \
> + "Found trace frame 0, tracepoint \[0-9\]+.
> +\#0 write_error_trace_file ().*" \
> + "tfind 0 on trace file bt.tf"
> +
> +gdb_test "backtrace" \
> +"\#0 write_error_trace_file ().*
> +\#1 .* in main .*" \
> + "backtrace on trace file bt.tf"
gdb_test_sequence is convenient to match the output of bt.
--
Yao (齐尧)
^ permalink raw reply [flat|nested] 8+ messages in thread* Re: [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read
2012-06-20 10:09 [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read Hui Zhu
2012-06-20 14:19 ` Yao Qi
@ 2012-06-20 14:44 ` Yao Qi
2012-06-21 11:16 ` Hui Zhu
1 sibling, 1 reply; 8+ messages in thread
From: Yao Qi @ 2012-06-20 14:44 UTC (permalink / raw)
To: gdb-patches; +Cc: Hui Zhu, Pedro Alves
Hi, Hui,
I got some fails with your patch applied,
FAIL: gdb.trace/tfile.exp: tstart
FAIL: gdb.trace/tfile.exp: tstop
FAIL: gdb.trace/tfile.exp: load trace file bt.tf (GDB internal error)
FAIL: gdb.trace/tfile.exp: tfind 0 on trace file bt.tf
FAIL: gdb.trace/tfile.exp: backtrace on trace file bt.tf
On Wednesday 20 June 2012 18:08:17 Hui Zhu wrote:
> gdb_test_no_output "tstart"
You start tracing in native, which doesn't support tracepoint.
> +gdb_test "break done_making_trace_files" ".*" ""
> +gdb_test "continue" ".*" ""
> +gdb_test_no_output "tstop"
> +gdb_test "tsave ./bt.tf" ".*" ""
AFAIK, in tfile.exp, trace file is generated by tfile.c and GDB only consumes
these trace files (basic.tf and error.tf) for test purpose.
--
Yao (齐尧)
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read
2012-06-20 14:44 ` Yao Qi
@ 2012-06-21 11:16 ` Hui Zhu
2012-06-21 14:17 ` Pedro Alves
0 siblings, 1 reply; 8+ messages in thread
From: Hui Zhu @ 2012-06-21 11:16 UTC (permalink / raw)
To: Yao Qi; +Cc: gdb-patches, Pedro Alves
On Wed, Jun 20, 2012 at 10:44 PM, Yao Qi <yao@codesourcery.com> wrote:
> Hi, Hui,
>
> I got some fails with your patch applied,
>
> FAIL: gdb.trace/tfile.exp: tstart
> FAIL: gdb.trace/tfile.exp: tstop
> FAIL: gdb.trace/tfile.exp: load trace file bt.tf (GDB internal error)
> FAIL: gdb.trace/tfile.exp: tfind 0 on trace file bt.tf
> FAIL: gdb.trace/tfile.exp: backtrace on trace file bt.tf
>
> On Wednesday 20 June 2012 18:08:17 Hui Zhu wrote:
>> gdb_test_no_output "tstart"
>
> You start tracing in native, which doesn't support tracepoint.
>
>> +gdb_test "break done_making_trace_files" ".*" ""
>> +gdb_test "continue" ".*" ""
>> +gdb_test_no_output "tstop"
>> +gdb_test "tsave ./bt.tf" ".*" ""
>
> AFAIK, in tfile.exp, trace file is generated by tfile.c and GDB only consumes
> these trace files (basic.tf and error.tf) for test purpose.
>
> --
> Yao (齐尧)
Thanks for your help.
Post a new version according to your comments.
Thanks,
Hui
2012-06-21 Hui Zhu <hui_zhu@mentor.com>
* gdb.trace/tfile.exp: Add new test for backtrace commands
with tfile to test the tfile memory read.
---
testsuite/gdb.trace/tfile.exp | 57 ++++++++++++++++++++++++++++++++++++++++++
1 file changed, 57 insertions(+)
--- a/testsuite/gdb.trace/tfile.exp
+++ b/testsuite/gdb.trace/tfile.exp
@@ -117,3 +117,60 @@ Not looking at any trace frame.*" \
gdb_test \
"interpreter-exec mi \"-target-select tfile basic.tf\"" \
"\\^connected.*"
+
+#Test the backtrace commands with tfile to test the tfile memory read
+if ![gdb_target_supports_trace] {
+ unsupported "target does not support trace"
+ return -1;
+}
+
+if ![is_remote target] {
+ unsupported "target is not remote"
+ return -1;
+}
+
+remote_file host delete bt.tf
+
+clean_restart ${testfile}
+
+runto_main
+
+set tp1 [gdb_gettpnum write_error_trace_file];
+
+if { [istarget i?86-*-*] } {
+ gdb_trace_setactions "Set actions for collect stack" \
+ "$tp1" \
+ "collect *(unsigned char *)\$esp@128" "^$"
+} elseif {[istarget x86_64-*-* ]} {
+ gdb_trace_setactions "Set actions for collect stack" \
+ "$tp1" \
+ "collect *(unsigned char *)\$rsp@128" "^$"
+} elseif {[istarget arm*-*-* ] || [istarget mips*-*-* ]} {
+ gdb_trace_setactions "Set actions for collect stack" \
+ "$tp1" \
+ "collect *(unsigned char *)\$sp@128" "^$"
+} else {
+ unsupported "don't know howto use tracepoint collect stack in this target"
+ continue
+}
+
+gdb_test_no_output "tstart"
+gdb_test "break done_making_trace_files" ".*" ""
+gdb_test "continue" ".*" ""
+gdb_test_no_output "tstop"
+gdb_test "tsave ./bt.tf" ".*" ""
+
+clean_restart ${testfile}
+
+gdb_test "target tfile bt.tf" "Created tracepoint.*" "load trace file bt.tf"
+gdb_test "tfind 0" \
+ "Found trace frame 0, tracepoint \[0-9\]+.
+\#0 write_error_trace_file ().*" \
+ "tfind 0 on trace file bt.tf"
+
+gdb_test_sequence "backtrace" "backtrace on trace file bt.tf" {
+ "\[\r\n\]+#0 .* write_error_trace_file \(\)"
+ "\[\r\n\]+#1 .* main"
+}
+
+remote_file host delete bt.tf
^ permalink raw reply [flat|nested] 8+ messages in thread* Re: [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read
2012-06-21 11:16 ` Hui Zhu
@ 2012-06-21 14:17 ` Pedro Alves
2012-06-22 1:43 ` Hui Zhu
0 siblings, 1 reply; 8+ messages in thread
From: Pedro Alves @ 2012-06-21 14:17 UTC (permalink / raw)
To: Hui Zhu; +Cc: Yao Qi, gdb-patches
On 06/21/2012 12:15 PM, Hui Zhu wrote:
>> >
>> > AFAIK, in tfile.exp, trace file is generated by tfile.c and GDB only consumes
>> > these trace files (basic.tf and error.tf) for test purpose.
>> >
> Thanks for your help.
> Post a new version according to your comments.
>
> Thanks,
> Hui
>
> 2012-06-21 Hui Zhu <hui_zhu@mentor.com>
>
> * gdb.trace/tfile.exp: Add new test for backtrace commands
> with tfile to test the tfile memory read.
How was this tested?
--
Pedro Alves
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read
2012-06-21 14:17 ` Pedro Alves
@ 2012-06-22 1:43 ` Hui Zhu
2012-06-22 9:05 ` Pedro Alves
0 siblings, 1 reply; 8+ messages in thread
From: Hui Zhu @ 2012-06-22 1:43 UTC (permalink / raw)
To: Pedro Alves; +Cc: Yao Qi, gdb-patches
On Thu, Jun 21, 2012 at 10:16 PM, Pedro Alves <palves@redhat.com> wrote:
> On 06/21/2012 12:15 PM, Hui Zhu wrote:
>
>>> >
>>> > AFAIK, in tfile.exp, trace file is generated by tfile.c and GDB only consumes
>>> > these trace files (basic.tf and error.tf) for test purpose.
>>> >
>> Thanks for your help.
>> Post a new version according to your comments.
>>
>> Thanks,
>> Hui
>>
>> 2012-06-21 Hui Zhu <hui_zhu@mentor.com>
>>
>> * gdb.trace/tfile.exp: Add new test for backtrace commands
>> with tfile to test the tfile memory read.
>
>
> How was this tested?
>
> --
> Pedro Alves
?
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read
2012-06-22 1:43 ` Hui Zhu
@ 2012-06-22 9:05 ` Pedro Alves
2012-06-24 7:39 ` Hui Zhu
0 siblings, 1 reply; 8+ messages in thread
From: Pedro Alves @ 2012-06-22 9:05 UTC (permalink / raw)
To: Hui Zhu; +Cc: Yao Qi, gdb-patches
On 06/22/2012 02:42 AM, Hui Zhu wrote:
> On Thu, Jun 21, 2012 at 10:16 PM, Pedro Alves <palves@redhat.com> wrote:
>> On 06/21/2012 12:15 PM, Hui Zhu wrote:
>>
>>>>>
>>>>> AFAIK, in tfile.exp, trace file is generated by tfile.c and GDB only consumes
>>>>> these trace files (basic.tf and error.tf) for test purpose.
>>>>>
>>> Thanks for your help.
>>> Post a new version according to your comments.
>>>
>>> Thanks,
>>> Hui
>>>
>>> 2012-06-21 Hui Zhu <hui_zhu@mentor.com>
>>>
>>> * gdb.trace/tfile.exp: Add new test for backtrace commands
>>> with tfile to test the tfile memory read.
>>
>>
>> How was this tested?
> ?
$ make check RUNTESTFLAGS="--target_board=native-gdbserver tfile.exp -v"
...
Skipping tfile.exp because of no fileio capabilities.
...
--
Pedro Alves
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read
2012-06-22 9:05 ` Pedro Alves
@ 2012-06-24 7:39 ` Hui Zhu
0 siblings, 0 replies; 8+ messages in thread
From: Hui Zhu @ 2012-06-24 7:39 UTC (permalink / raw)
To: Pedro Alves; +Cc: Yao Qi, gdb-patches
On Fri, Jun 22, 2012 at 5:05 PM, Pedro Alves <palves@redhat.com> wrote:
> On 06/22/2012 02:42 AM, Hui Zhu wrote:
>
>> On Thu, Jun 21, 2012 at 10:16 PM, Pedro Alves <palves@redhat.com> wrote:
>>> On 06/21/2012 12:15 PM, Hui Zhu wrote:
>>>
>>>>>>
>>>>>> AFAIK, in tfile.exp, trace file is generated by tfile.c and GDB only consumes
>>>>>> these trace files (basic.tf and error.tf) for test purpose.
>>>>>>
>>>> Thanks for your help.
>>>> Post a new version according to your comments.
>>>>
>>>> Thanks,
>>>> Hui
>>>>
>>>> 2012-06-21 Hui Zhu <hui_zhu@mentor.com>
>>>>
>>>> * gdb.trace/tfile.exp: Add new test for backtrace commands
>>>> with tfile to test the tfile memory read.
>>>
>>>
>>> How was this tested?
>
>
>> ?
>
>
> $ make check RUNTESTFLAGS="--target_board=native-gdbserver tfile.exp -v"
> ...
> Skipping tfile.exp because of no fileio capabilities.
> ...
>
> --
> Pedro Alves
This is because:
if [target_info exists gdb,nofileio] {
verbose "Skipping tfile.exp because of no fileio capabilities."
continue
}
So to use this test, need remove "set_board_info gdb,do_reload_on_run 1"
Thanks,
Hui
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2012-06-24 7:39 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2012-06-20 10:09 [PATCH] add testsuite for for backtrace commands with tfile to test the tfile memory read Hui Zhu
2012-06-20 14:19 ` Yao Qi
2012-06-20 14:44 ` Yao Qi
2012-06-21 11:16 ` Hui Zhu
2012-06-21 14:17 ` Pedro Alves
2012-06-22 1:43 ` Hui Zhu
2012-06-22 9:05 ` Pedro Alves
2012-06-24 7:39 ` Hui Zhu
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox