Re: [RFC] Re: Parsing kernel parameters and escaping "

[Daniel Mierswa <impulze@impulze.org>]

[-- Attachment #1.1: Type: text/plain, Size: 335 bytes --]

Second version of the patch, I accidently forgot to take care of out of
bounds checking when reaching the end of the kernel parameter line.

-- 
Mierswa, Daniel

If you still don't like it, that's ok: that's why I'm boss. I simply
know better than you do.
               --- Linus Torvalds, comp.os.linux.advocacy, 1996/07/22

[-- Attachment #1.2: kernel.patch --]
[-- Type: text/plain, Size: 4265 bytes --]

From fe3bf00a0fc2db941fef41c44fc6d9eefad4d037 Mon Sep 17 00:00:00 2001
From: Daniel Mierswa <impulze@impulze.org>
Date: Tue, 7 Jul 2009 00:54:38 +0200
Subject: [PATCH] Attempt to handle quotes in kernel parameters

There was a limitation for kernel parameters with regards to quoting. It
wasn't possible to escape quotes or use quotes to form space-filled
values _inside_ parameters. This patch attempts to make that possible,
kernel parameters are now parsed as follows:
'"param= value"' [param= value][]
'param=" value "" combination "' [param][ value  combination ]
'param=" \" test"' [param][ " test]
'"param"=another' [param][another]
---
 kernel/params.c |  146 +++++++++++++++++++++++++++++++++++++++----------------
 1 files changed, 104 insertions(+), 42 deletions(-)

diff --git a/kernel/params.c b/kernel/params.c
index de273ec..73de8f5 100644
--- a/kernel/params.c
+++ b/kernel/params.c
@@ -75,58 +75,120 @@ static int parse_one(char *param,
 	return -ENOENT;
 }
 
-/* You can use " around spaces, but can't escape ". */
-/* Hyphens and underscores equivalent in parameter names. */
-static char *next_arg(char *args, char **param, char **val)
+/* handle quotes in tokens (parameter and values)
+ * '" foo bar "' => ' foo bar '
+ * '" foo \" "'  => ' foo " '
+ */
+static void add_token(char ** token, char * args)
 {
-	unsigned int i, equals = 0;
-	int in_quote = 0, quoted = 0;
-	char *next;
+	char * iterator, * last_quote;
+	int in_quotes;
 
-	if (*args == '"') {
-		args++;
-		in_quote = 1;
-		quoted = 1;
+	in_quotes = 0;
+	last_quote = NULL;
+
+	for (iterator = args; *iterator; iterator++) {
+		if (*iterator == '\\' && *(iterator + 1) == '"') {
+			char * mover;
+
+			/* move all characters back */
+			for (mover = iterator; *mover; mover++) {
+				*mover = *(mover + 1);
+			}
+
+			continue;
+		}
+
+		if (*iterator != '"') {
+			continue;
+		}
+
+		if (in_quotes) {
+			char * mover;
+
+			/* move whole string back until current " is reached */
+			for (mover = last_quote; mover != iterator - 1; mover++) {
+				*mover = *(mover + 1);
+			}
+
+			/* ignore the current " and move the rest of the string back */
+			while (*mover) {
+				*mover = *(mover + 2);
+				mover++;
+			}
+
+			/* ignored 2 quotes, decrease the iterator */
+			iterator -= 2;
+			last_quote = NULL;
+		} else {
+			last_quote = iterator;
+		}
+
+		in_quotes = !in_quotes;
 	}
 
-	for (i = 0; args[i]; i++) {
-		if (args[i] == ' ' && !in_quote)
+	*token = args;
+}
+
+static char * next_arg(char * args, char ** param, char ** val)
+{
+	char * token, * next;
+	int in_quotes, is_escaped;
+
+	*param = *val = next = NULL;
+	in_quotes = is_escaped = 0;
+
+	/* tokenizer */
+	for (token = args; *token; token++) {
+		/* parameter or value is finished */
+		if (!in_quotes && *token == ' ') {
+			*token = '\0';
+
+			/* there're still characters in the parameter line */
+			next = token + 1;
+
+			/* remove trailing whitespace */
+			while (*next == ' ') {
+				next++;
+			}
+
 			break;
-		if (equals == 0) {
-			if (args[i] == '=')
-				equals = i;
 		}
-		if (args[i] == '"')
-			in_quote = !in_quote;
-	}
 
-	*param = args;
-	if (!equals)
-		*val = NULL;
-	else {
-		args[equals] = '\0';
-		*val = args + equals + 1;
-
-		/* Don't include quotes in value. */
-		if (**val == '"') {
-			(*val)++;
-			if (args[i-1] == '"')
-				args[i-1] = '\0';
+		/* parameter/value split */
+		if (!in_quotes && *token == '=' && !*param) {
+			*token = '\0';
+			add_token(param, args);
+			args = token + 1;
+			continue;
 		}
-		if (quoted && args[i-1] == '"')
-			args[i-1] = '\0';
+
+		if (!is_escaped && *token == '\\') {
+			is_escaped = 1;
+			continue;
+		}
+
+		if (*token == '"' && !is_escaped) {
+			in_quotes = !in_quotes;
+		}
+
+		/* always reset escape value, only " needs it */
+		is_escaped = 0;
 	}
 
-	if (args[i]) {
-		args[i] = '\0';
-		next = args + i + 1;
-	} else
-		next = args + i;
+	if (!*param) {
+		add_token(param, args);
+	} else {
+		add_token(val, args);
+	}
+
+	/* there're parameters left in the command line */
+	if (next) {
+		return next;
+	}
 
-	/* Chew up trailing spaces. */
-	while (*next == ' ')
-		next++;
-	return next;
+	/* end of the line */
+	return token;
 }
 
 /* Args looks like "foo=bar,bar2 baz=fuz wiz". */
-- 
1.6.3.3


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 198 bytes --]