Received-SPF: pass (google.com: domain of linux-kernel+bounces-150792-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45e3:2400::1 as permitted sender) client-ip=2604:1380:45e3:2400::1;
Precedence: bulk
MIME-Version: 1.0
References: <20240416061533.921723-1-irogers@google.com> <20240416061533.921723-12-irogers@google.com>
 <ac8835f8-0ea5-4f28-941c-aa43f0da92fd@linux.intel.com>
In-Reply-To: <ac8835f8-0ea5-4f28-941c-aa43f0da92fd@linux.intel.com>
From: Ian Rogers <irogers@google.com>
Date: Thu, 18 Apr 2024 14:07:31 -0700
Message-ID: <CAP-5=fXkn-nFTqGEqBYt6NUvoXU7OyLJeSCnWdD3taLHyK-xtQ@mail.gmail.com>
Subject: Re: [PATCH v2 11/16] perf parse-events: Improve error message for bad numbers
To: "Liang, Kan" <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@redhat.com>, 
	Arnaldo Carvalho de Melo <acme@kernel.org>, Namhyung Kim <namhyung@kernel.org>, 
	Mark Rutland <mark.rutland@arm.com>, 
	Alexander Shishkin <alexander.shishkin@linux.intel.com>, Jiri Olsa <jolsa@kernel.org>, 
	Adrian Hunter <adrian.hunter@intel.com>, James Clark <james.clark@arm.com>, 
	linux-perf-users@vger.kernel.org, linux-kernel@vger.kernel.org, 
	bpf@vger.kernel.org, Atish Patra <atishp@rivosinc.com>, linux-riscv@lists.infradead.org, 
	Beeman Strong <beeman@rivosinc.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Thu, Apr 18, 2024 at 1:27=E2=80=AFPM Liang, Kan <kan.liang@linux.intel.c=
om> wrote:
>
>
>
> On 2024-04-16 2:15 a.m., Ian Rogers wrote:
> > Use the error handler from the parse_state to give a more informative
> > error message.
> >
> > Before:
> > ```
> > $ perf stat -e 'cycles/period=3D99999999999999999999/' true
> > event syntax error: 'cycles/period=3D99999999999999999999/'
> >                                   \___ parser error
> > Run 'perf list' for a list of valid events
> >
> >  Usage: perf stat [<options>] [<command>]
> >
> >     -e, --event <event>   event selector. use 'perf list' to list avail=
able events
> > ```
> >
> > After:
> > ```
> > $ perf stat -e 'cycles/period=3D99999999999999999999/' true
> > event syntax error: 'cycles/period=3D99999999999999999999/'
> >                                   \___ parser error
> >
> > event syntax error: '..les/period=3D99999999999999999999/'
> >                                   \___ Bad base 10 number "999999999999=
99999999"
>
>
> It seems the patch only works for decimal?
>
> ./perf stat -e 'cycles/period=3D0xaaaaaaaaaaaaaaaaaaaaaa/' true
> event syntax error: '..les/period=3D0xaaaaaaaaaaaaaaaaaaaaaa/'
>                                    \___ parser error
>  Run 'perf list' for a list of valid events
>
>   Usage: perf stat [<options>] [<command>]
>
>      -e, --event <event>   event selector. use 'perf list' to list
> available events
>
> Thanks,
> Kan

Right, for hexadecimal we say the number of digits is at most 16, so
when you exceed this the token is no longer recognized. It just
becomes input that can't be parsed, hence parser error. Doing this
means we can simplify other strtoull checks, but I agree having a
better error message for hexadecimal would be good. Let's do it as
follow up.

Thanks,
Ian

> > Run 'perf list' for a list of valid events
> >
> >  Usage: perf stat [<options>] [<command>]
> >
> >     -e, --event <event>   event selector. use 'perf list' to list avail=
able events
> > ```
> >
> > Signed-off-by: Ian Rogers <irogers@google.com>
> > ---
> >  tools/perf/util/parse-events.l | 40 ++++++++++++++++++++--------------
> >  1 file changed, 24 insertions(+), 16 deletions(-)
> >
> > diff --git a/tools/perf/util/parse-events.l b/tools/perf/util/parse-eve=
nts.l
> > index 6fe37003ab7b..0cd68c9f0d4f 100644
> > --- a/tools/perf/util/parse-events.l
> > +++ b/tools/perf/util/parse-events.l
> > @@ -18,26 +18,34 @@
> >
> >  char *parse_events_get_text(yyscan_t yyscanner);
> >  YYSTYPE *parse_events_get_lval(yyscan_t yyscanner);
> > +int parse_events_get_column(yyscan_t yyscanner);
> > +int parse_events_get_leng(yyscan_t yyscanner);
> >
> > -static int __value(YYSTYPE *yylval, char *str, int base, int token)
> > +static int get_column(yyscan_t scanner)
> >  {
> > -     u64 num;
> > -
> > -     errno =3D 0;
> > -     num =3D strtoull(str, NULL, base);
> > -     if (errno)
> > -             return PE_ERROR;
> > -
> > -     yylval->num =3D num;
> > -     return token;
> > +     return parse_events_get_column(scanner) - parse_events_get_leng(s=
canner);
> >  }
> >
> > -static int value(yyscan_t scanner, int base)
> > +static int value(struct parse_events_state *parse_state, yyscan_t scan=
ner, int base)
> >  {
> >       YYSTYPE *yylval =3D parse_events_get_lval(scanner);
> >       char *text =3D parse_events_get_text(scanner);
> > +     u64 num;
> >
> > -     return __value(yylval, text, base, PE_VALUE);
> > +     errno =3D 0;
> > +     num =3D strtoull(text, NULL, base);
> > +     if (errno) {
> > +             struct parse_events_error *error =3D parse_state->error;
> > +             char *help =3D NULL;
> > +
> > +             if (asprintf(&help, "Bad base %d number \"%s\"", base, te=
xt) > 0)
> > +                     parse_events_error__handle(error, get_column(scan=
ner), help , NULL);
> > +
> > +             return PE_ERROR;
> > +     }
> > +
> > +     yylval->num =3D num;
> > +     return PE_VALUE;
> >  }
> >
> >  static int str(yyscan_t scanner, int token)
> > @@ -283,8 +291,8 @@ r0x{num_raw_hex}  { return str(yyscanner, PE_RAW); =
}
> >        */
> >  "/"/{digit}          { return PE_BP_SLASH; }
> >  "/"/{non_digit}              { BEGIN(config); return '/'; }
> > -{num_dec}            { return value(yyscanner, 10); }
> > -{num_hex}            { return value(yyscanner, 16); }
> > +{num_dec}            { return value(_parse_state, yyscanner, 10); }
> > +{num_hex}            { return value(_parse_state, yyscanner, 16); }
> >       /*
> >        * We need to separate 'mem:' scanner part, in order to get speci=
fic
> >        * modifier bits parsed out. Otherwise we would need to handle PE=
_NAME
> > @@ -330,8 +338,8 @@ cgroup-switches                                   {=
 return sym(yyscanner, PERF_COUNT_SW_CGROUP_SWITCHES); }
> >  {lc_type}-{lc_op_result}-{lc_op_result}      { return str(yyscanner, P=
E_LEGACY_CACHE); }
> >  mem:                 { BEGIN(mem); return PE_PREFIX_MEM; }
> >  r{num_raw_hex}               { return str(yyscanner, PE_RAW); }
> > -{num_dec}            { return value(yyscanner, 10); }
> > -{num_hex}            { return value(yyscanner, 16); }
> > +{num_dec}            { return value(_parse_state, yyscanner, 10); }
> > +{num_hex}            { return value(_parse_state, yyscanner, 16); }
> >
> >  {modifier_event}     { return str(yyscanner, PE_MODIFIER_EVENT); }
> >  {name}                       { return str(yyscanner, PE_NAME); }