qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Frederic Konrad <fred.konrad@greensocs.com>
To: "lig.fnst@cn.fujitsu.com" <lig.fnst@cn.fujitsu.com>,
	"qemu-devel@nongnu.org" <qemu-devel@nongnu.org>
Cc: "peter.maydell@linaro.org" <peter.maydell@linaro.org>,
	"quintela@redhat.com" <quintela@redhat.com>,
	"mark.burton@greensocs.com" <mark.burton@greensocs.com>,
	"dgilbert@redhat.com" <dgilbert@redhat.com>,
	"amit.shah@redhat.com" <amit.shah@redhat.com>,
	"pbonzini@redhat.com" <pbonzini@redhat.com>,
	"vilanova@ac.upc.edu" <vilanova@ac.upc.edu>
Subject: Re: [Qemu-devel] 答复:  [RFC PATCH v5 10/13] gdbstub: allow reverse execution in gdb stub.
Date: Thu, 26 Jun 2014 09:39:46 +0200	[thread overview]
Message-ID: <53ABCE42.2020704@greensocs.com> (raw)
In-Reply-To: <534C5107B281854290643468BBAAD41F17293E@G08CNEXMBPEKD01.g08.fujitsu.local>

On 26/06/2014 04:50, lig.fnst@cn.fujitsu.com wrote:
> Hi, 
>
> Why not a gdb record support?
>
> Thanks!
> Li Guang
Hi,

Good idea.
We didn't think about that yet.
It might be done in a futur patch-set.

Thanks,
Fred

>> -----邮件原件-----
>> 发件人: qemu-devel-bounces+lig.fnst=cn.fujitsu.com@nongnu.org
>> [mailto:qemu-devel-bounces+lig.fnst=cn.fujitsu.com@nongnu.org] 代表
>> fred.konrad@greensocs.com
>> 发送时间: 2014年6月25日 23:20
>> 收件人: qemu-devel@nongnu.org
>> 抄送: peter.maydell@linaro.org; quintela@redhat.com;
>> mark.burton@greensocs.com; dgilbert@redhat.com; amit.shah@redhat.com;
>> pbonzini@redhat.com; vilanova@ac.upc.edu; fred.konrad@greensocs.com
>> 主题: [Qemu-devel] [RFC PATCH v5 10/13] gdbstub: allow reverse execution in
>> gdb stub.
>>
>> From: KONRAD Frederic <fred.konrad@greensocs.com>
>>
>> This allows gdb to reverse step QEMU: reverse-stepi and reverse-cont
>> commands are allowed.
>>
>> When step_backward is called, QEMU restores a snapshot before the actual
>> instruction and stops (with a debug exit) when the previous instruction is
>> reached.
>>
>> Signed-off-by: KONRAD Frederic <fred.konrad@greensocs.com>
>> ---
>>  gdbstub.c | 31 ++++++++++++++++++++++++++++++-
>>  1 file changed, 30 insertions(+), 1 deletion(-)
>>
>> diff --git a/gdbstub.c b/gdbstub.c
>> index 8afe0b7..99769af 100644
>> --- a/gdbstub.c
>> +++ b/gdbstub.c
>> @@ -33,6 +33,7 @@
>>  #include "sysemu/char.h"
>>  #include "sysemu/sysemu.h"
>>  #include "exec/gdbstub.h"
>> +#include "reverse-execution.h"
>>  #endif
>>
>>  #define MAX_PACKET_LENGTH 4096
>> @@ -1113,6 +1114,17 @@ static int gdb_handle_packet(GDBState *s, const
>> char *line_buf)
>>              if (cc->gdb_core_xml_file != NULL) {
>>                  pstrcat(buf, sizeof(buf), ";qXfer:features:read+");
>>              }
>> +
>> +            #ifndef CONFIG_USER_ONLY
>> +            /*
>> +             * When reverse execution is enabled those additional
>> features must
>> +             * be set so GDB allows reverse-stepi and reverse-continue
>> command.
>> +             */
>> +            if (cexe_is_enabled()) {
>> +                pstrcat(buf, sizeof(buf),
>> ";ReverseStep+;ReverseContinue+");
>> +            }
>> +            #endif /* !CONFIG_USER_ONLY */
>> +
>>              put_packet(s, buf);
>>              break;
>>          }
>> @@ -1161,7 +1173,23 @@ static int gdb_handle_packet(GDBState *s, const
>> char *line_buf)
>>          }
>>          /* Unrecognised 'q' command.  */
>>          goto unknown_command;
>> -
>> +    #ifndef CONFIG_USER_ONLY
>> +    case 'b':
>> +        /* Reverse execution. */
>> +        switch (*p) {
>> +        case 's':
>> +            cexe_step_backward(s->c_cpu, 1);
>> +            break;
>> +        case 'c':
>> +            cexe_continue_backward(s->c_cpu);
>> +            break;
>> +        default:
>> +            buf[0] = '\0';
>> +            put_packet(s, buf);
>> +            break;
>> +        }
>> +        break;
>> +    #endif /* !CONFIG_USER_ONLY */
>>      default:
>>      unknown_command:
>>          /* put empty packet */
>> @@ -1221,6 +1249,7 @@ static void gdb_vm_state_change(void *opaque, int
>> running, RunState state)
>>          ret = GDB_SIGNAL_TRAP;
>>          break;
>>      case RUN_STATE_PAUSED:
>> +        cexe_stop_reverse_continue();
>>          ret = GDB_SIGNAL_INT;
>>          break;
>>      case RUN_STATE_SHUTDOWN:
>> --
>> 1.9.0
>>

  parent reply	other threads:[~2014-06-26  7:39 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-06-25 15:19 [Qemu-devel] [RFC PATCH v5 00/13] Reverse execution fred.konrad
2014-06-25 15:19 ` [Qemu-devel] [RFC PATCH v5 01/13] icount: put icount variables into TimerState fred.konrad
2014-06-25 15:19 ` [Qemu-devel] [RFC PATCH v5 02/13] migration: migrate icount fields fred.konrad
2014-06-25 17:38   ` Juan Quintela
2014-06-25 15:19 ` [Qemu-devel] [RFC PATCH v5 03/13] migration: make qemu_savevm_state public fred.konrad
2014-06-25 15:19 ` [Qemu-devel] [RFC PATCH v5 04/13] icount: introduce icount timer fred.konrad
2014-06-25 15:19 ` [Qemu-devel] [RFC PATCH v5 05/13] icount: check for icount clock deadline when cpu loop exits fred.konrad
2014-06-25 15:20 ` [Qemu-devel] [RFC PATCH v5 06/13] icount: make icount extra computed on icount clock as well fred.konrad
2014-06-25 15:20 ` [Qemu-devel] [RFC PATCH v5 07/13] timer: add cpu_icount_to_ns function fred.konrad
2014-06-25 15:20 ` [Qemu-devel] [RFC PATCH v5 08/13] trace-events: add reverse-execution events fred.konrad
2014-06-25 15:20 ` [Qemu-devel] [RFC PATCH v5 09/13] introduce reverse execution mechanism fred.konrad
     [not found]   ` <534C5107B281854290643468BBAAD41F172901@G08CNEXMBPEKD01.g08.fujitsu.local>
2014-06-26  7:28     ` [Qemu-devel] 答复: " Frederic Konrad
     [not found]       ` <534C5107B281854290643468BBAAD41F172C78@G08CNEXMBPEKD01.g08.fujitsu.local>
2014-06-26  8:20         ` [Qemu-devel] 答复: " Frederic Konrad
2014-06-25 15:20 ` [Qemu-devel] [RFC PATCH v5 10/13] gdbstub: allow reverse execution in gdb stub fred.konrad
     [not found]   ` <534C5107B281854290643468BBAAD41F17293E@G08CNEXMBPEKD01.g08.fujitsu.local>
2014-06-26  7:39     ` Frederic Konrad [this message]
2014-06-25 15:20 ` [Qemu-devel] [RFC PATCH v5 11/13] cpu-exec: trigger a debug request when rexec stops fred.konrad
2014-06-25 15:20 ` [Qemu-devel] [RFC PATCH v5 12/13] cexe: synchronize icount on the next event fred.konrad
2014-06-25 15:20 ` [Qemu-devel] [RFC PATCH v5 13/13] cexe: allow to enable reverse execution fred.konrad
2014-06-25 15:43 ` [Qemu-devel] [RFC PATCH v5 00/13] Reverse execution Paolo Bonzini
2014-06-26  8:17   ` Frederic Konrad
2014-06-26 10:14     ` Paolo Bonzini
2014-06-26 13:55       ` Frederic Konrad
2014-06-26 13:56         ` Paolo Bonzini
2014-06-26 15:11           ` Frederic Konrad
2014-06-26 15:32             ` Paolo Bonzini
2014-06-26 15:52               ` Sebastian Tanase
2014-06-27  8:08                 ` Frederic Konrad

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=53ABCE42.2020704@greensocs.com \
    --to=fred.konrad@greensocs.com \
    --cc=amit.shah@redhat.com \
    --cc=dgilbert@redhat.com \
    --cc=lig.fnst@cn.fujitsu.com \
    --cc=mark.burton@greensocs.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=quintela@redhat.com \
    --cc=vilanova@ac.upc.edu \
    /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 NNTP newsgroup(s).