From: Al Viro <viro@zeniv.linux.org.uk>
To: linux-sparse@vger.kernel.org
Cc: chriscli@google.com, torvalds@linux-foundation.org,
zxh@xh-zhang.com, ben.dooks@codethink.co.uk,
dan.carpenter@linaro.org, rf@opensource.cirrus.com
Subject: [PATCH 09/21] steal 2 bits from argnum for argument kind
Date: Mon, 16 Mar 2026 07:04:03 +0000 [thread overview]
Message-ID: <20260316070415.768839-9-viro@zeniv.linux.org.uk> (raw)
In-Reply-To: <20260316070415.768839-1-viro@zeniv.linux.org.uk>
We have 3 separate token types (TOKEN_{MACRO,QUOTED,STR}_ARGUMENT),
with fairly similar handling at expansion time. Let's steal two bits
from ->argnum and use them to represent the kind of occurrence; that
simplifies substitute() and allows for better code generation there.
The object we use to store the argument state at expansion time (struct
arg) is already a structure with 3 pointers to token lists (unexpanded,
expanded and stringified forms of the argument) and 3 integer counters -
the number of remaining occurrencies of each kind. Gather those into
3-element arrays indexed by the kind; counts will be gone soon, token
lists will remain.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
---
pre-process.c | 106 +++++++++++++++++++++-----------------------------
token.h | 14 +++++--
2 files changed, 55 insertions(+), 65 deletions(-)
diff --git a/pre-process.c b/pre-process.c
index efb208e7..b45688d5 100644
--- a/pre-process.c
+++ b/pre-process.c
@@ -304,12 +304,8 @@ static struct token *collect_arg(struct token *prev, bool vararg, const struct p
*/
struct arg {
- struct token *arg;
- struct token *expanded;
- struct token *str;
- int n_normal;
- int n_quoted;
- int n_str;
+ struct token *arg[3];
+ int count[3];
};
static int collect_arguments(struct token *start, struct symbol *sym, struct arg *args, struct token *what)
@@ -330,10 +326,10 @@ static int collect_arguments(struct token *start, struct symbol *sym, struct arg
goto Eclosing;
p = &arglist->next->count;
arglist = arglist->next->next;
- args[commas].arg = start->next;
- args[commas].n_normal = p->normal;
- args[commas].n_quoted = p->quoted;
- args[commas].n_str = p->str;
+ args[commas].arg[ARG_QUOTED] = start->next;
+ args[commas].count[ARG_NORMAL] = p->normal;
+ args[commas].count[ARG_QUOTED] = p->quoted;
+ args[commas].count[ARG_STR] = p->str;
if (!match_op(next, ',')) {
if (commas < fixed - 1)
goto Efew;
@@ -353,10 +349,10 @@ static int collect_arguments(struct token *start, struct symbol *sym, struct arg
goto Eexcess;
if (vararg) {
p = &arglist->next->count;
- args[fixed].arg = v;
- args[fixed].n_normal = p->normal;
- args[fixed].n_quoted = p->quoted;
- args[fixed].n_str = p->str;
+ args[fixed].arg[ARG_QUOTED] = v;
+ args[fixed].count[ARG_NORMAL] = p->normal;
+ args[fixed].count[ARG_QUOTED] = p->quoted;
+ args[fixed].count[ARG_STR] = p->str;
}
what->next = next->next;
return 1;
@@ -440,21 +436,21 @@ static void expand_arguments(int count, struct arg *args)
{
int i;
for (i = 0; i < count; i++) {
- struct token *arg = args[i].arg;
+ struct token *arg = args[i].arg[ARG_QUOTED];
if (!arg)
arg = &eof_token_entry;
- if (args[i].n_str)
- args[i].str = stringify(arg);
- if (args[i].n_normal) {
- if (!args[i].n_quoted) {
- args[i].expanded = arg;
- args[i].arg = NULL;
+ if (args[i].count[ARG_STR])
+ args[i].arg[ARG_STR] = stringify(arg);
+ if (args[i].count[ARG_NORMAL]) {
+ if (!args[i].count[ARG_QUOTED]) {
+ args[i].arg[ARG_NORMAL] = arg;
+ args[i].arg[ARG_QUOTED] = NULL;
} else if (eof_token(arg)) {
- args[i].expanded = arg;
+ args[i].arg[ARG_NORMAL] = arg;
} else {
- args[i].expanded = dup_list(arg);
+ args[i].arg[ARG_NORMAL] = dup_list(arg);
}
- expand_list(&args[i].expanded);
+ expand_list(&args[i].arg[ARG_NORMAL]);
}
}
}
@@ -629,13 +625,18 @@ static inline int argnum(const struct token *arg)
return arg->argnum >> ARGNUM_BITS_STOLEN;
}
+static inline enum arg_kind argkind(const struct token *arg)
+{
+ return arg->argnum & ARGNUM_KIND_MASK;
+}
+
static int handle_kludge(const struct token **p, struct arg *args)
{
const struct token *t = (*p)->next->next;
while (1) {
- struct arg *v = &args[argnum(t)];
+ struct token *v = args[argnum(t)].arg[ARG_QUOTED];
if (token_type(t->next) != TOKEN_CONCAT) {
- if (v->arg) {
+ if (v) {
/* ignore the first ## */
*p = (*p)->next;
return 0;
@@ -644,7 +645,7 @@ static int handle_kludge(const struct token **p, struct arg *args)
*p = t;
return 1;
}
- if (v->arg && !eof_token(v->arg))
+ if (v && !eof_token(v))
return 0; /* no magic */
t = t->next->next;
}
@@ -685,14 +686,9 @@ static struct token **substitute(struct token **list, const struct token *body,
tail = &added->next;
break;
- case TOKEN_STR_ARGUMENT:
- arg = args[argnum(body)].str;
- count = &args[argnum(body)].n_str;
- goto copy_arg;
-
- case TOKEN_QUOTED_ARGUMENT:
- arg = args[argnum(body)].arg;
- count = &args[argnum(body)].n_quoted;
+ case TOKEN_MACRO_ARGUMENT:
+ arg = args[argnum(body)].arg[argkind(body)];
+ count = &args[argnum(body)].count[argkind(body)];
if (!arg || eof_token(arg)) {
if (state == Concat)
state = Normal;
@@ -700,16 +696,6 @@ static struct token **substitute(struct token **list, const struct token *body,
state = Placeholder;
continue;
}
- goto copy_arg;
-
- case TOKEN_MACRO_ARGUMENT:
- arg = args[argnum(body)].expanded;
- count = &args[argnum(body)].n_normal;
- if (eof_token(arg)) {
- state = Normal;
- continue;
- }
- copy_arg:
if (!--*count)
tail = move_into(&added, arg);
else
@@ -1040,8 +1026,6 @@ static int token_different(struct token *t1, struct token *t2)
different = t1->special != t2->special;
break;
case TOKEN_MACRO_ARGUMENT:
- case TOKEN_QUOTED_ARGUMENT:
- case TOKEN_STR_ARGUMENT:
different = t1->argnum != t2->argnum;
break;
case TOKEN_CHAR_EMBEDDED_0 ... TOKEN_CHAR_EMBEDDED_3:
@@ -1206,7 +1190,7 @@ Eva_args:
return NULL;
}
-static int try_arg(struct token *token, enum token_type type, struct token *arglist)
+static int try_arg(struct token *token, enum arg_kind kind, struct token *arglist)
{
struct ident *ident = token->ident;
int nr, n;
@@ -1224,13 +1208,13 @@ static int try_arg(struct token *token, enum token_type type, struct token *argl
for (int i = 0; i < nr; i++)
arglist = arglist->next->next;
- token->argnum = nr << ARGNUM_BITS_STOLEN;
- token_type(token) = type;
- switch (type) {
- case TOKEN_MACRO_ARGUMENT:
+ token->argnum = (nr << ARGNUM_BITS_STOLEN) | kind;
+ token_type(token) = TOKEN_MACRO_ARGUMENT;
+ switch (kind) {
+ case ARG_NORMAL:
n = ++arglist->next->count.normal;
break;
- case TOKEN_QUOTED_ARGUMENT:
+ case ARG_QUOTED:
n = ++arglist->next->count.quoted;
break;
default:
@@ -1251,7 +1235,7 @@ static struct token *handle_hash(struct token **p, struct token *arglist)
struct token *token = *p;
if (macro_funclike) {
struct token *next = token->next;
- if (!try_arg(next, TOKEN_STR_ARGUMENT, arglist))
+ if (!try_arg(next, ARG_STR, arglist))
goto Equote;
next->pos.whitespace = token->pos.whitespace;
__free_token(token);
@@ -1273,7 +1257,7 @@ static struct token *handle_hashhash(struct token *token, struct token *arglist)
struct token *concat;
int state = match_op(token, ',');
- try_arg(token, TOKEN_QUOTED_ARGUMENT, arglist);
+ try_arg(token, ARG_QUOTED, arglist);
while (1) {
struct token *t;
@@ -1297,7 +1281,7 @@ static struct token *handle_hashhash(struct token *token, struct token *arglist)
return NULL;
}
- is_arg = try_arg(t, TOKEN_QUOTED_ARGUMENT, arglist);
+ is_arg = try_arg(t, ARG_QUOTED, arglist);
if (state == 1 && is_arg) {
state = is_arg;
@@ -1339,7 +1323,7 @@ static struct token *parse_expansion(struct token *expansion, struct token *argl
if (!token)
return NULL;
} else {
- try_arg(token, TOKEN_MACRO_ARGUMENT, arglist);
+ try_arg(token, ARG_NORMAL, arglist);
}
if (token_type(token) == TOKEN_ERROR)
goto Earg;
@@ -2007,7 +1991,7 @@ static int handle_nondirective(struct stream *stream, struct token **line, struc
static struct token *first_arg(struct arg *args)
{
- struct token *arg = args[0].arg;
+ struct token *arg = args[0].arg[ARG_QUOTED];
expand_list(&arg);
return arg;
}
@@ -2356,11 +2340,9 @@ static void dump_macro(struct symbol *sym)
case TOKEN_CONCAT:
printf("##");
break;
- case TOKEN_STR_ARGUMENT:
- printf("#");
- /* fall-through */
- case TOKEN_QUOTED_ARGUMENT:
case TOKEN_MACRO_ARGUMENT:
+ if (argkind(token) == ARG_STR)
+ printf("#");
printf("%s", show_ident(args[argnum(token)]));
break;
default:
diff --git a/token.h b/token.h
index fe7c7fe9..273da39a 100644
--- a/token.h
+++ b/token.h
@@ -100,8 +100,6 @@ enum token_type {
TOKEN_STREAMBEGIN,
TOKEN_STREAMEND,
TOKEN_MACRO_ARGUMENT,
- TOKEN_STR_ARGUMENT,
- TOKEN_QUOTED_ARGUMENT,
TOKEN_CONCAT,
TOKEN_GNU_KLUDGE,
TOKEN_UNTAINT,
@@ -177,8 +175,18 @@ struct argcount {
unsigned str:10;
};
+enum arg_kind {
+ ARG_QUOTED = 0,
+ ARG_NORMAL = 1,
+ ARG_STR = 2,
+};
+
+enum {
+ ARGNUM_BITS_STOLEN = 2
+};
+
enum {
- ARGNUM_BITS_STOLEN
+ ARGNUM_KIND_MASK = 3
};
/*
--
2.47.3
next prev parent reply other threads:[~2026-03-16 7:01 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <cover.1771930766.git.dan.carpenter@linaro.org>
2026-02-24 11:07 ` [PATCH] sparse: add support for __VA_OPT__ Dan Carpenter
2026-02-24 11:16 ` Ben Dooks
2026-02-24 11:56 ` Dan Carpenter
2026-02-24 12:42 ` Richard Fitzgerald
2026-02-24 13:15 ` Ben Dooks
2026-02-25 2:39 ` Chris Li
2026-02-25 3:36 ` Al Viro
2026-02-25 5:29 ` [RFC PATCH] pre-process: add __VA_OPT__ support Eric Zhang
2026-02-25 6:40 ` Al Viro
2026-02-25 7:27 ` Al Viro
2026-02-25 8:14 ` Eric Zhang
2026-02-25 22:18 ` Al Viro
2026-02-26 7:29 ` Al Viro
2026-03-16 6:56 ` Al Viro
2026-03-16 7:03 ` [PATCH 01/21] split copy() into "need to copy" and "can move in place" cases Al Viro
2026-03-16 7:03 ` [PATCH 02/21] expand and simplify the call of dup_token() in copy() Al Viro
2026-03-16 7:03 ` [PATCH 03/21] more dup_token() optimizations Al Viro
2026-03-16 7:03 ` [PATCH 04/21] parsing #define: saner handling of argument count, part 1 Al Viro
2026-03-16 7:03 ` [PATCH 05/21] simplify collect_arguments() and fix error handling there Al Viro
2026-03-16 7:04 ` [PATCH 06/21] try_arg(): don't use arglist for argument name lookups Al Viro
2026-03-16 7:04 ` [PATCH 07/21] make expand_has_...() responsible for expanding its argument Al Viro
2026-03-16 7:04 ` [PATCH 08/21] preparing to change argument number encoding for TOKEN_..._ARGUMENT Al Viro
2026-03-16 7:04 ` Al Viro [this message]
2026-03-16 7:04 ` [PATCH 10/21] on-demand argument expansion Al Viro
2026-03-16 7:04 ` [PATCH 11/21] kill create_arglist() Al Viro
2026-03-16 7:04 ` [PATCH 12/21] stop mangling arglist, get rid of TOKEN_ARG_COUNT Al Viro
2026-03-16 7:04 ` [PATCH 13/21] deal with ## on arguments separately Al Viro
2026-03-16 7:04 ` [PATCH 14/21] preparations for __VA_OPT__ support: reshuffle argument slot assignments Al Viro
2026-03-16 7:04 ` [PATCH 15/21] pre-process.c: split try_arg() Al Viro
2026-03-16 7:04 ` [PATCH 16/21] __VA_OPT__: parsing Al Viro
2026-03-16 7:04 ` [PATCH 17/21] expansion-time va_opt handling Al Viro
2026-03-16 7:04 ` [PATCH 18/21] merge(): saner handling of ->noexpand Al Viro
2026-03-16 7:04 ` [PATCH 19/21] simplify the calling conventions of collect_arguments() Al Viro
2026-03-16 7:04 ` [PATCH 20/21] make expand_one_symbol() inline Al Viro
2026-03-16 7:04 ` [PATCH 21/21] substitute(): convert switch() into cascade of ifs Al Viro
2026-03-16 16:42 ` [RFC PATCH] pre-process: add __VA_OPT__ support Linus Torvalds
2026-03-19 3:53 ` Al Viro
2026-03-19 4:07 ` Linus Torvalds
2026-03-19 5:34 ` Al Viro
2026-03-17 7:41 ` Chris Li
2026-03-18 6:35 ` Eric Zhang
2026-02-25 7:05 ` [PATCH] sparse: add support for __VA_OPT__ Chris Li
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=20260316070415.768839-9-viro@zeniv.linux.org.uk \
--to=viro@zeniv.linux.org.uk \
--cc=ben.dooks@codethink.co.uk \
--cc=chriscli@google.com \
--cc=dan.carpenter@linaro.org \
--cc=linux-sparse@vger.kernel.org \
--cc=rf@opensource.cirrus.com \
--cc=torvalds@linux-foundation.org \
--cc=zxh@xh-zhang.com \
/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