Hi,
Recently, the distributor has set "utf8" to NLS_DEFAULT, therefore,
FAT uses the "iocharset=utf8" as default. But, since "iocharset=utf8"
doesn't provide the function (lower <-> upper conversion) which FAT
needs, so FAT can't provide suitable behavior.
Then, this patch does,
- doesn't recognize "utf8" as "iocharset"
- doesn't use NLS_DEFAULT as default "iocharset"
- instead of NLS_DEFAULT, adds FAT_DEFAULT_CODEPAGE and
FAT_DEFAULT_IOCHARSET
NOTE: the following looks like buggy, so it's not recommended
"codepage=437,iocharset=iso8859-1,utf8"
however, some utf8 file name can handle. (in this case, it uses the
table of iso8859-1 for lower <-> upper conversion)
Sign-off-by: Jesse Barnes <[email protected]>
Sign-off-by: OGAWA Hirofumi <[email protected]>
---
fs/Kconfig | 19 ++++++++++++
fs/fat/inode.c | 88 +++++++++++++++++++++++++++++----------------------------
2 files changed, 65 insertions(+), 42 deletions(-)
diff -puN fs/Kconfig~fat_default-nls fs/Kconfig
--- linux-2.6.7/fs/Kconfig~fat_default-nls 2004-06-20 03:30:39.000000000 +0900
+++ linux-2.6.7-hirofumi/fs/Kconfig 2004-06-20 03:30:39.000000000 +0900
@@ -672,6 +672,25 @@ config VFAT_FS
To compile this as a module, choose M here: the module will be called
vfat.
+config FAT_DEFAULT_CODEPAGE
+ int "Default codepage for FAT"
+ depends on MSDOS_FS || VFAT_FS
+ default 437
+ help
+ This option should be set to the codepage of your FAT filesystems.
+ It can be overridden with the 'codepage' mount option.
+
+config FAT_DEFAULT_IOCHARSET
+ string "Default iocharset for FAT"
+ depends on VFAT_FS
+ default "iso8859-1"
+ help
+ Set this to the default I/O character set you'd like FAT to use.
+ It should probably match the character set that most of your
+ FAT filesystems use, and can be overridded with the 'iocharset'
+ mount option for FAT filesystems. Note that UTF8 is *not* a
+ supported charset for FAT filesystems.
+
config UMSDOS_FS
#dep_tristate ' UMSDOS: Unix-like file system on top of standard MSDOS fs' CONFIG_UMSDOS_FS $CONFIG_MSDOS_FS
# UMSDOS is temprory broken
diff -puN fs/fat/inode.c~fat_default-nls fs/fat/inode.c
--- linux-2.6.7/fs/fat/inode.c~fat_default-nls 2004-06-20 03:30:39.000000000 +0900
+++ linux-2.6.7-hirofumi/fs/fat/inode.c 2004-06-20 03:30:39.000000000 +0900
@@ -23,6 +23,14 @@
#include <linux/parser.h>
#include <asm/unaligned.h>
+#ifndef CONFIG_FAT_DEFAULT_IOCHARSET
+/* if user don't select VFAT, this is undefined. */
+#define CONFIG_FAT_DEFAULT_IOCHARSET ""
+#endif
+
+static int fat_default_codepage = CONFIG_FAT_DEFAULT_CODEPAGE;
+static char fat_default_iocharset[] = CONFIG_FAT_DEFAULT_IOCHARSET;
+
/*
* New FAT inode stuff. We do the following:
* a) i_ino is constant and has nothing with on-disk location.
@@ -166,20 +174,17 @@ void fat_put_super(struct super_block *s
if (sbi->nls_disk) {
unload_nls(sbi->nls_disk);
sbi->nls_disk = NULL;
- sbi->options.codepage = 0;
+ sbi->options.codepage = fat_default_codepage;
}
if (sbi->nls_io) {
unload_nls(sbi->nls_io);
sbi->nls_io = NULL;
}
- /*
- * Note: the iocharset option might have been specified
- * without enabling nls_io, so check for it here.
- */
- if (sbi->options.iocharset) {
+ if (sbi->options.iocharset != fat_default_iocharset) {
kfree(sbi->options.iocharset);
- sbi->options.iocharset = NULL;
+ sbi->options.iocharset = fat_default_iocharset;
}
+
sb->s_fs_info = NULL;
kfree(sbi);
}
@@ -196,11 +201,11 @@ static int fat_show_options(struct seq_f
seq_printf(m, ",gid=%u", opts->fs_gid);
seq_printf(m, ",fmask=%04o", opts->fs_fmask);
seq_printf(m, ",dmask=%04o", opts->fs_dmask);
- if (sbi->nls_disk)
+ if (sbi->nls_disk && opts->codepage != fat_default_codepage)
seq_printf(m, ",codepage=%s", sbi->nls_disk->charset);
if (isvfat) {
- if (sbi->nls_io
- && strcmp(sbi->nls_io->charset, CONFIG_NLS_DEFAULT))
+ if (sbi->nls_io &&
+ strcmp(opts->iocharset, fat_default_iocharset))
seq_printf(m, ",iocharset=%s", sbi->nls_io->charset);
switch (opts->shortname) {
@@ -331,14 +336,15 @@ static int parse_options(char *options,
char *p;
substring_t args[MAX_OPT_ARGS];
int option;
+ char *iocharset;
opts->isvfat = is_vfat;
opts->fs_uid = current->uid;
opts->fs_gid = current->gid;
opts->fs_fmask = opts->fs_dmask = current->fs->umask;
- opts->codepage = 0;
- opts->iocharset = NULL;
+ opts->codepage = fat_default_codepage;
+ opts->iocharset = fat_default_iocharset;
if (is_vfat)
opts->shortname = VFAT_SFN_DISPLAY_LOWER|VFAT_SFN_CREATE_WIN95;
else
@@ -351,7 +357,7 @@ static int parse_options(char *options,
*debug = 0;
if (!options)
- return 1;
+ return 0;
while ((p = strsep(&options, ",")) != NULL) {
int token;
@@ -437,10 +443,12 @@ static int parse_options(char *options,
/* vfat specific */
case Opt_charset:
- kfree(opts->iocharset);
- opts->iocharset = match_strdup(&args[0]);
- if (!opts->iocharset)
- return 0;
+ if (opts->iocharset != fat_default_iocharset)
+ kfree(opts->iocharset);
+ iocharset = match_strdup(&args[0]);
+ if (!iocharset)
+ return -ENOMEM;
+ opts->iocharset = iocharset;
break;
case Opt_shortname_lower:
opts->shortname = VFAT_SFN_DISPLAY_LOWER
@@ -486,14 +494,20 @@ static int parse_options(char *options,
default:
printk(KERN_ERR "FAT: Unrecognized mount option \"%s\" "
"or missing value\n", p);
- return 0;
+ return -EINVAL;
}
}
+ /* UTF8 doesn't provide FAT semantics */
+ if (!strcmp(opts->iocharset, "utf8")) {
+ printk(KERN_ERR "FAT: utf8 is not a valid IO charset"
+ " for FAT filesystems\n");
+ return -EINVAL;
+ }
if (opts->unicode_xlate)
opts->utf8 = 0;
- return 1;
+ return 0;
}
static int fat_calc_dir_size(struct inode *inode)
@@ -784,7 +798,7 @@ int fat_fill_super(struct super_block *s
struct msdos_sb_info *sbi;
u16 logical_sector_size;
u32 total_sectors, total_clusters, fat_clusters, rootdir_sectors;
- int debug, cp, first;
+ int debug, first;
unsigned int media;
long error;
char buf[50];
@@ -801,8 +815,8 @@ int fat_fill_super(struct super_block *s
sb->s_export_op = &fat_export_ops;
sbi->dir_ops = fs_dir_inode_ops;
- error = -EINVAL;
- if (!parse_options(data, isvfat, &debug, &sbi->options))
+ error = parse_options(data, isvfat, &debug, &sbi->options);
+ if (error)
goto out_fail;
fat_cache_init(sb);
@@ -1009,31 +1023,21 @@ int fat_fill_super(struct super_block *s
}
error = -EINVAL;
- cp = sbi->options.codepage ? sbi->options.codepage : 437;
- sprintf(buf, "cp%d", cp);
+ sprintf(buf, "cp%d", sbi->options.codepage);
sbi->nls_disk = load_nls(buf);
if (!sbi->nls_disk) {
- /* Fail only if explicit charset specified */
- if (sbi->options.codepage != 0) {
- printk(KERN_ERR "FAT: codepage %s not found\n", buf);
- goto out_fail;
- }
- sbi->options.codepage = 0; /* already 0?? */
- sbi->nls_disk = load_nls_default();
+ printk(KERN_ERR "FAT: codepage %s not found\n", buf);
+ goto out_fail;
}
/* FIXME: utf8 is using iocharset for upper/lower conversion */
if (sbi->options.isvfat) {
- if (sbi->options.iocharset != NULL) {
- sbi->nls_io = load_nls(sbi->options.iocharset);
- if (!sbi->nls_io) {
- printk(KERN_ERR
- "FAT: IO charset %s not found\n",
- sbi->options.iocharset);
- goto out_fail;
- }
- } else
- sbi->nls_io = load_nls_default();
+ sbi->nls_io = load_nls(sbi->options.iocharset);
+ if (!sbi->nls_io) {
+ printk(KERN_ERR "FAT: IO charset %s not found\n",
+ sbi->options.iocharset);
+ goto out_fail;
+ }
}
error = -ENOMEM;
@@ -1068,7 +1072,7 @@ out_fail:
unload_nls(sbi->nls_io);
if (sbi->nls_disk)
unload_nls(sbi->nls_disk);
- if (sbi->options.iocharset)
+ if (sbi->options.iocharset != fat_default_iocharset)
kfree(sbi->options.iocharset);
sb->s_fs_info = NULL;
kfree(sbi);
_
--
OGAWA Hirofumi <[email protected]>
#include <hallo.h>
* OGAWA Hirofumi [Sun, Jun 20 2004, 11:41:51PM]:
> Recently, the distributor has set "utf8" to NLS_DEFAULT, therefore,
Excuse my ignorance but I will comment this having things from various
HOWTOs in mind.
> FAT uses the "iocharset=utf8" as default. But, since "iocharset=utf8"
> doesn't provide the function (lower <-> upper conversion) which FAT
> needs, so FAT can't provide suitable behavior.
I (as user) am completely confused now. I though that "iocharset" was
the option to convert the (UCS) VFAT names from Unicode (on disk) to some
limited encoding represented the Unix system (eg. ISO-8859-1) and vice
versa; but it was limited to single-byte encodings, so "utf8" switch was
invented to do direct UCS<->UTF-8 conversion.
> Then, this patch does,
>
> - doesn't recognize "utf8" as "iocharset"
> - doesn't use NLS_DEFAULT as default "iocharset"
Is it possible to extend iocharset to allow UCS<->UTF-8 conversion?
Wouldn't this make the utf8 switch obsolete?
> - instead of NLS_DEFAULT, adds FAT_DEFAULT_CODEPAGE and
> FAT_DEFAULT_IOCHARSET
Excuse me, why do you need a special IOCHARSET setting for FAT
filesystems? Why should it be different from any other filesystems? Are
there any other filesystems except those from MS that support (widechar)
Unicode filenames on disk?
> NOTE: the following looks like buggy, so it's not recommended
>
> "codepage=437,iocharset=iso8859-1,utf8"
>
> however, some utf8 file name can handle. (in this case, it uses the
> table of iso8859-1 for lower <-> upper conversion)
What do you man with "lower <-> upper"? The letter case? Isn't this a
task for the FAT16-names access method?
> +config FAT_DEFAULT_CODEPAGE
> + int "Default codepage for FAT"
> + depends on MSDOS_FS || VFAT_FS
> + default 437
> + help
> + This option should be set to the codepage of your FAT filesystems.
> + It can be overridden with the 'codepage' mount option.
Should describe explicitely that only the short "MS-DOS" names are meant
here.
> +config FAT_DEFAULT_IOCHARSET
> + string "Default iocharset for FAT"
> + depends on VFAT_FS
> + default "iso8859-1"
> + help
> + Set this to the default I/O character set you'd like FAT to use.
What is the "I/O charset"? The description is completely useless for a
user, it does not explain what is converted and where and why and what
happens with the data or what is this good for. In contrary, I would
think that this is the conversion method from a local encoding to equaly
limited I/O encoding.
> + It should probably match the character set that most of your
> + FAT filesystems use, and can be overridded with the 'iocharset'
> + mount option for FAT filesystems. Note that UTF8 is *not* a
> + supported charset for FAT filesystems.
UTF-8 is not a charset, it is an encoding. Don't mix things together.
Regards,
Eduard.
--
Die ganz ganz Schlauen sehen um f?nf Ecken - und sind geradeaus blind.
-- Benjamin Franklin
Eduard Bloch <[email protected]> writes:
> > FAT uses the "iocharset=utf8" as default. But, since "iocharset=utf8"
> > doesn't provide the function (lower <-> upper conversion) which FAT
> > needs, so FAT can't provide suitable behavior.
>
> I (as user) am completely confused now. I though that "iocharset" was
> the option to convert the (UCS) VFAT names from Unicode (on disk) to some
> limited encoding represented the Unix system (eg. ISO-8859-1) and vice
> versa;
No, iocharset is used for user visible filename.
Is the following enough as example?
user
^ |
| |
<euc-jp> <euc-jp>
| |
------ lookup ---------------- create ------
| |
<euc-jp> <euc-jp>
| |
["iocharset"] ["iocharset"]
| |
<UCS> <UCS>
| |
+-----+-----+ (long name)
| | | |
["codepage"] | ["codepage"] |
| | | |
(8.3 alias) (long name) (8.3 alias) |
^ ^ | |
| | V V
<shift-jis> <UCS> <shift-jis> <UCS>
------------------------------------------
disk
> but it was limited to single-byte encodings, so "utf8" switch was
> invented to do direct UCS<->UTF-8 conversion.
No. But multi-byte encodings has some problem on the letter case conversion.
> > Then, this patch does,
> >
> > - doesn't recognize "utf8" as "iocharset"
> > - doesn't use NLS_DEFAULT as default "iocharset"
>
> Is it possible to extend iocharset to allow UCS<->UTF-8 conversion?
> Wouldn't this make the utf8 switch obsolete?
Yes. But I think that we'll need to rewrite the NLS stuff and some
part of filesystems.
> > - instead of NLS_DEFAULT, adds FAT_DEFAULT_CODEPAGE and
> > FAT_DEFAULT_IOCHARSET
>
> Excuse me, why do you need a special IOCHARSET setting for FAT
> filesystems? Why should it be different from any other filesystems? Are
> there any other filesystems except those from MS that support (widechar)
> Unicode filenames on disk?
NLS_DEFAULT is used by befs, cifs, vfat, isofs, ncpfs, udf.
> > NOTE: the following looks like buggy, so it's not recommended
> >
> > "codepage=437,iocharset=iso8859-1,utf8"
> >
> > however, some utf8 file name can handle. (in this case, it uses the
> > table of iso8859-1 for lower <-> upper conversion)
>
> What do you man with "lower <-> upper"? The letter case?
Yes. Sorry for my bad English.
> Isn't this a task for the FAT16-names access method?
Yes, but 8.3-alias is stored always even if it's long name.
> > +config FAT_DEFAULT_CODEPAGE
> > + int "Default codepage for FAT"
> > + depends on MSDOS_FS || VFAT_FS
> > + default 437
> > + help
> > + This option should be set to the codepage of your FAT filesystems.
> > + It can be overridden with the 'codepage' mount option.
>
> Should describe explicitely that only the short "MS-DOS" names are meant
> here.
Ok. Any suggestion?
> > +config FAT_DEFAULT_IOCHARSET
> > + string "Default iocharset for FAT"
> > + depends on VFAT_FS
> > + default "iso8859-1"
> > + help
> > + Set this to the default I/O character set you'd like FAT to use.
>
> What is the "I/O charset"? The description is completely useless for a
> user, it does not explain what is converted and where and why and what
> happens with the data or what is this good for. In contrary, I would
> think that this is the conversion method from a local encoding to equaly
> limited I/O encoding.
Yes, but it's very old, is widely used in linux filesystems.
If it changed, old users may be confused instead.
> > + It should probably match the character set that most of your
> > + FAT filesystems use, and can be overridded with the 'iocharset'
> > + mount option for FAT filesystems. Note that UTF8 is *not* a
> > + supported charset for FAT filesystems.
>
> UTF-8 is not a charset, it is an encoding. Don't mix things together.
Yes, I'll fix. How about the following?
> > + supported encoding for FAT filesystems.
Thanks
--
OGAWA Hirofumi <[email protected]>
OGAWA Hirofumi <[email protected]> writes:
> > > NOTE: the following looks like buggy, so it's not recommended
> > >
> > > "codepage=437,iocharset=iso8859-1,utf8"
> > >
> > > however, some utf8 file name can handle. (in this case, it uses the
> > > table of iso8859-1 for lower <-> upper conversion)
>
> > Isn't this a task for the FAT16-names access method?
>
> Yes, but 8.3-alias is stored always even if it's long name.
And of course table is also used for case insensible access.
--
OGAWA Hirofumi <[email protected]>