linux-kernel - Re: [PATCH] modpost: simplify modpost

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CAK7LNARscy9jcJwnxFEqYFkKekzkN1Sw8d7XMkT1y3a3EGpjHg@mail.gmail.com>
Date: Fri, 16 Aug 2024 22:44:03 +0900
From: Masahiro Yamada <masahiroy@...nel.org>
To: Nathan Chancellor <nathan@...nel.org>
Cc: linux-kbuild@...r.kernel.org, linux-kernel@...r.kernel.org, 
	Nicolas Schier <nicolas@...sle.eu>
Subject: Re: [PATCH] modpost: simplify modpost_log()

On Tue, Aug 13, 2024 at 6:23 AM Nathan Chancellor <nathan@...nel.org> wrote:
>
> Hi Masahiro,
>
> On Mon, Aug 12, 2024 at 11:45:39PM +0900, Masahiro Yamada wrote:
> > With commit cda5f94e88b4 ("modpost: avoid using the alias attribute"),
> > only two log levels remain: LOG_WARN and LOG_ERROR. Simplify this by
> > making it a boolean variable.
> >
> > Signed-off-by: Masahiro Yamada <masahiroy@...nel.org>
> > ---
> >
> >  scripts/mod/modpost.c | 17 ++++++-----------
> >  scripts/mod/modpost.h | 11 +++--------
> >  2 files changed, 9 insertions(+), 19 deletions(-)
> >
> > diff --git a/scripts/mod/modpost.c b/scripts/mod/modpost.c
> > index d0f138803207..c896872862dc 100644
> > --- a/scripts/mod/modpost.c
> > +++ b/scripts/mod/modpost.c
> > @@ -67,20 +67,15 @@ static unsigned int nr_unresolved;
> >
> >  #define MODULE_NAME_LEN (64 - sizeof(Elf_Addr))
> >
> > -void modpost_log(enum loglevel loglevel, const char *fmt, ...)
> > +void modpost_log(bool is_error, const char *fmt, ...)
> >  {
> >       va_list arglist;
> >
> > -     switch (loglevel) {
> > -     case LOG_WARN:
> > -             fprintf(stderr, "WARNING: ");
> > -             break;
> > -     case LOG_ERROR:
> > +     if (is_error) {
> >               fprintf(stderr, "ERROR: ");
> >               error_occurred = true;
> > -             break;
> > -     default: /* invalid loglevel, ignore */
> > -             break;
> > +     } else {
> > +             fprintf(stderr, "WARNING: ");
> >       }
> >
> >       fprintf(stderr, "modpost: ");
> > @@ -1692,7 +1687,7 @@ static void check_exports(struct module *mod)
> >               exp = find_symbol(s->name);
> >               if (!exp) {
> >                       if (!s->weak && nr_unresolved++ < MAX_UNRESOLVED_REPORTS)
> > -                             modpost_log(warn_unresolved ? LOG_WARN : LOG_ERROR,
> > +                             modpost_log(warn_unresolved,
>
> I think the logic is reversed here? If warn_unresolved was true before,
> LOG_WARN would be used but I think warn_unresolved being true with this
> change would trigger the behavior of LOG_ERROR. Should it be
> '!warn_unresolved'?


Thanks, I will fix this in v2.




>
> >                                           "\"%s\" [%s.ko] undefined!\n",
> >                                           s->name, mod->name);
> >                       continue;
> > @@ -1715,7 +1710,7 @@ static void check_exports(struct module *mod)
> >                       basename = mod->name;
> >
> >               if (!contains_namespace(&mod->imported_namespaces, exp->namespace)) {
> > -                     modpost_log(allow_missing_ns_imports ? LOG_WARN : LOG_ERROR,
> > +                     modpost_log(allow_missing_ns_imports,
>
> Same thing here?
>
> Alternatively, I suppose the logic could be reversed in modpost_log()
> and is_error could be turned into no_error or something?
>
> >                                   "module %s uses symbol %s from namespace %s, but does not import it.\n",
> >                                   basename, exp->name, exp->namespace);
> >                       add_namespace(&mod->missing_namespaces, exp->namespace);
> > diff --git a/scripts/mod/modpost.h b/scripts/mod/modpost.h
> > index f756e6578b9e..6f418f0afd04 100644
> > --- a/scripts/mod/modpost.h
> > +++ b/scripts/mod/modpost.h
> > @@ -184,13 +184,8 @@ char *read_text_file(const char *filename);
> >  char *get_line(char **stringp);
> >  void *sym_get_data(const struct elf_info *info, const Elf_Sym *sym);
> >
> > -enum loglevel {
> > -     LOG_WARN,
> > -     LOG_ERROR,
> > -};
> > -
> >  void __attribute__((format(printf, 2, 3)))
> > -modpost_log(enum loglevel loglevel, const char *fmt, ...);
> > +modpost_log(bool is_error, const char *fmt, ...);
> >
> >  /*
> >   * warn - show the given message, then let modpost continue running, still
> > @@ -205,6 +200,6 @@ modpost_log(enum loglevel loglevel, const char *fmt, ...);
> >   * fatal - show the given message, and bail out immediately. This should be
> >   *         used when there is no point to continue running modpost.
> >   */
> > -#define warn(fmt, args...)   modpost_log(LOG_WARN, fmt, ##args)
> > -#define error(fmt, args...)  modpost_log(LOG_ERROR, fmt, ##args)
> > +#define warn(fmt, args...)   modpost_log(false, fmt, ##args)
> > +#define error(fmt, args...)  modpost_log(true, fmt, ##args)
>
> I suppose the declaration of modpost_log() is close enough to see what
> the true/false argument means but maybe something like
>
>   modpost_log(/* is_error = */ false, ...)
>
> could be nice? No strong opinion here though, feel free to just ignore
> this whole comment entirely if you disagree!
>
> >  #define fatal(fmt, args...)  do { error(fmt, ##args); exit(1); } while (1)
> > --
> > 2.43.0
> >



-- 
Best Regards
Masahiro Yamada