public inbox for gdb-patches@sourceware.org
 help / color / mirror / Atom feed
From: Bruno Larsen <blarsen@redhat.com>
To: "gdb-patches@sourceware.org" <gdb-patches@sourceware.org>
Subject: [PINGv2] [PATCH] gdb/reverse: Fix stepping over recursive functions
Date: Wed, 15 Jun 2022 16:21:22 -0300	[thread overview]
Message-ID: <e85df77c-5d3b-1be6-c94b-8dc491becce3@redhat.com> (raw)
In-Reply-To: <badbbb6c-8b2a-7eec-1f46-4f59ede90d39@redhat.com>

ping!

Cheers!
Bruno Larsen

On 6/8/22 08:17, Bruno Larsen wrote:
> Ping!
> 
> Cheers!
> Bruno Larsen
> 
> On 5/25/22 15:02, Bruno Larsen wrote:
>> Currently, when using GDB to do reverse debugging, if we try to use the
>> command "reverse next" to skip a recursive function, instead of skipping
>> all of the recursive calls and stopping in the previous line, we stop at
>> the second to last recursive call, and need to manually step backwards
>> until we leave the first call.  This is well documented in PR gdb/16678.
>>
>> This bug happens because when GDB notices that a reverse step has
>> entered into a function, GDB will add a step_resume_breakpoint at the
>> start of the function, then single step out of the prologue once that
>> breakpoint is hit.  Recursion poses a problem because that breakpoint will
>> be hit many times before GDB should actually stop the inferior.  To fix
>> this issue, when the step_resume_breakpoint is hit (and GDB is executing
>> backwards), we analyze if the caller of the frame is the original frame
>> where we started the "reverse next", and if it is, GDB will stop the
>> inferior, otherwise GDB will just ignore the breakpoint.
>>
>> This commit also changes gdb.reverse/step-reverse.exp to contain a
>> recursive function and attempt to both, skip it altogether, and to skip
>> the second call from inside the first call, as this setup broke a
>> previous version of the patch.
>> ---
>>   gdb/infrun.c                               | 24 +++++++++
>>   gdb/testsuite/gdb.reverse/step-reverse.c   | 16 +++++-
>>   gdb/testsuite/gdb.reverse/step-reverse.exp | 59 ++++++++++++++++++++--
>>   3 files changed, 95 insertions(+), 4 deletions(-)
>>
>> diff --git a/gdb/infrun.c b/gdb/infrun.c
>> index 02c98b50c8c..95fb3227aa3 100644
>> --- a/gdb/infrun.c
>> +++ b/gdb/infrun.c
>> @@ -6761,6 +6761,30 @@ process_event_stop_test (struct execution_control_state *ecs)
>>       case BPSTAT_WHAT_STEP_RESUME:
>>         infrun_debug_printf ("BPSTAT_WHAT_STEP_RESUME");
>> +      /* If we are executing backwards, we are doing a "next" and the
>> +     current frame is not the same as where we started, the
>> +     step_resume_breakpoint we have just hit has been added to the start
>> +     of a function so we can skip the whle function.  However, if we are
>> +     skipping a recursive call, we only want to act as if we hit the
>> +     breakpoint only when the caller of the current frame is the original
>> +     frame we were single stepping from.  */
>> +      if (execution_direction == EXEC_REVERSE
>> +      && ecs->event_thread->control.step_over_calls == STEP_OVER_ALL
>> +      && !frame_id_eq (get_stack_frame_id (get_current_frame ()),
>> +               ecs->event_thread->control.step_stack_frame_id))
>> +    {
>> +      infrun_debug_printf ("We are not done with recursing yet");
>> +      /* If the caller's ID is not the same as the starting frame, we
>> +         can ignore this breakpoint.  */
>> +      if (!frame_id_eq (frame_unwind_caller_id (get_current_frame ()),
>> +                ecs->event_thread->control.step_stack_frame_id))
>> +        {
>> +          keep_going (ecs);
>> +          return;
>> +        }
>> +    } else
>> +        infrun_debug_printf ("are we not recursing?");
>> +
>>         delete_step_resume_breakpoint (ecs->event_thread);
>>         if (ecs->event_thread->control.proceed_to_finish
>>         && execution_direction == EXEC_REVERSE)
>> diff --git a/gdb/testsuite/gdb.reverse/step-reverse.c b/gdb/testsuite/gdb.reverse/step-reverse.c
>> index aea2a98541d..a390ac2580c 100644
>> --- a/gdb/testsuite/gdb.reverse/step-reverse.c
>> +++ b/gdb/testsuite/gdb.reverse/step-reverse.c
>> @@ -26,6 +26,17 @@ int callee() {        /* ENTER CALLEE */
>>     return myglob++;    /* ARRIVED IN CALLEE */
>>   }            /* RETURN FROM CALLEE */
>> +/* We need to make this function take more than a single instruction
>> +   to run, otherwise it could hide PR gdb/16678, as reverse execution can
>> +   step over a single-instruction function.  */
>> +int recursive_callee (int val){
>> +    if (val == 0) return 0;
>> +    val /= 2;
>> +    if (val>1)
>> +    val++;
>> +    return recursive_callee (val);    /* RECURSIVE CALL */
>> +} /* EXIT RECURSIVE FUNCTION */
>> +
>>   /* A structure which, we hope, will need to be passed using memcpy.  */
>>   struct rhomboidal {
>>     int rather_large[100];
>> @@ -51,6 +62,9 @@ int main () {
>>      y = y + 4;
>>      z = z + 5;    /* STEP TEST 2 */
>> +   /* Test that next goes over recursive calls too */
>> +   recursive_callee (32); /* NEXT OVER THIS RECURSION */
>> +
>>      /* Test that "next" goes over a call */
>>      callee();    /* NEXT OVER THIS CALL */
>> @@ -60,7 +74,7 @@ int main () {
>>      /* Test "stepi" */
>>      a[5] = a[3] - a[4]; /* FINISH TEST */
>>      callee();    /* STEPI TEST */
>> -
>> +
>>      /* Test "nexti" */
>>      callee();    /* NEXTI TEST */
>> diff --git a/gdb/testsuite/gdb.reverse/step-reverse.exp b/gdb/testsuite/gdb.reverse/step-reverse.exp
>> index 997b62604d5..4f56b4785ca 100644
>> --- a/gdb/testsuite/gdb.reverse/step-reverse.exp
>> +++ b/gdb/testsuite/gdb.reverse/step-reverse.exp
>> @@ -47,9 +47,11 @@ gdb_test "step" ".*STEP TEST 1.*" "step test 1"
>>   gdb_test "next 2" ".*NEXT TEST 2.*" "next test 2"
>>   gdb_test "step 3" ".*STEP TEST 2.*" "step test 2"
>> +# next through a recursive function call
>> +gdb_test "next 2" "NEXT OVER THIS CALL.*" "next over recursion"
>> +
>>   # step over call
>> -gdb_test "step" ".*NEXT OVER THIS CALL.*" "step up to call"
>>   gdb_test "next" ".*STEP INTO THIS CALL.*" "next over call"
>>   # step into call
>> @@ -118,7 +120,7 @@ gdb_test_multiple "stepi" "$test_message" {
>>   set test_message "stepi back from function call"
>>   gdb_test_multiple "stepi" "$test_message" {
>> -    -re "NEXTI TEST.*$gdb_prompt $" {
>> +    -re -wrap "NEXTI TEST.*" {
>>       pass "$test_message"
>>       }
>>       -re "ARRIVED IN CALLEE.*$gdb_prompt $" {
>> @@ -143,7 +145,6 @@ gdb_test_multiple "stepi" "$test_message" {
>>   ###
>>   # Set reverse execution direction
>> -
>>   gdb_test_no_output "set exec-dir reverse" "set reverse execution"
>>   # stepi backward thru return and into a function
>> @@ -247,6 +248,58 @@ gdb_test_multiple "step" "$test_message" {
>>   gdb_test "next" ".*NEXT OVER THIS CALL.*" "reverse next over call"
>> +# Dont reverse the execution direction yet, as we will need another
>> +# forward step after this test
>> +
>> +set step_out 0
>> +gdb_test_multiple "next" "reverse next over recursion" {
>> +    -re -wrap ".*NEXT OVER THIS RECURSION.*" {
>> +    pass "reverse next over recursion"
>> +    }
>> +    -re -wrap ".*RECURSIVE CALL.*" {
>> +    fail "reverse next over recursion"
>> +    set step_out 1
>> +    }
>> +}
>> +if { "$step_out" == 1 } {
>> +    gdb_test_multiple "next" "stepping out of recursion" {
>> +    -re -wrap "NEXT OVER THIS RECURSION.*" {
>> +        set step_out 0
>> +    }
>> +    -re -wrap ".*" {
>> +        send_gdb "reverse-next\n"
>> +        exp_continue
>> +    }
>> +    }
>> +}
>> +
>> +# Step forward over recursion again so we can test stepping over calls
>> +# inside the recursion itself.
>> +gdb_test_no_output "set exec-dir forward" "forward again to test recursion"
>> +gdb_test "next" "NEXT OVER THIS CALL.*" "reverse next over recursion again"
>> +gdb_test_no_output "set exec-dir reverse" "reverse again to test recursion"
>> +
>> +gdb_test "step" ".*EXIT RECURSIVE FUNCTION.*" "enter recursive function"
>> +set step_pass 1
>> +gdb_test_multiple "next" "step over recursion inside the recursion" {
>> +    -re -wrap ".*EXIT RECURSIVE FUNCTION.*" {
>> +    set step_pass 0
>> +    send_gdb "next\n"
>> +    exp_continue
>> +    }
>> +    -re -wrap ".*NEXT OVER THIS RECURSION.*" {
>> +    if {$step_pass} {
>> +        pass "step over recursion inside the recursion"
>> +    } else {
>> +        fail "step over recursion inside the recursion"
>> +    }
>> +    }
>> +    -re -wrap ".*" {
>> +    send_gdb "next\n"
>> +    exp_continue
>> +    }
>> +}
>> +
>>   # step/next backward with count
>>   gdb_test "step 3" ".*REVERSE STEP TEST 1.*" "reverse step test 1"


  reply	other threads:[~2022-06-15 19:21 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-25 18:02 Bruno Larsen
2022-06-08 11:17 ` [PING] " Bruno Larsen
2022-06-15 19:21   ` Bruno Larsen [this message]
2022-06-22 18:49     ` [PINGv3] " Bruno Larsen
2022-06-29 18:46       ` [PINGv4] " Bruno Larsen
2022-07-06 11:47         ` [PINGv5] " Bruno Larsen
2022-07-07 18:48 ` Pedro Alves
2022-07-08 17:16   ` Bruno Larsen
2022-07-22 15:31   ` Bruno Larsen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=e85df77c-5d3b-1be6-c94b-8dc491becce3@redhat.com \
    --to=blarsen@redhat.com \
    --cc=gdb-patches@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).