JFFS2 uncompressed?

JFFS2 uncompressed?

[Steven Scholz]

Hi there,

IIRC files are stored compressed on a JFFS2 partition.

Is there any chance that I could use JFFS2 without any compression?

I guess this might me faster for our embedded system!

Thanks,

Steven

Re: JFFS2 uncompressed?

[Jörn Engel]

On Tue, 8 April 2003 15:59:35 +0200, Steven Scholz wrote:
> 
> IIRC files are stored compressed on a JFFS2 partition.
> 
> Is there any chance that I could use JFFS2 without any compression?
> 
> I guess this might me faster for our embedded system!

Piece of cake.
Look at fs/jffs2/compr.c, first function. Add a couple of #if 0 and
you're done.

Do the same with the second function and the kernel image should
shrink a little.

Jörn

-- 
Beware of bugs in the above code; I have only proved it correct, but
not tried it.
-- Donald Knuth

Re: JFFS2 uncompressed?

[Steven Scholz]

Jörn Engel schrieb:
> On Tue, 8 April 2003 15:59:35 +0200, Steven Scholz wrote:
> 
>>IIRC files are stored compressed on a JFFS2 partition.
>>
>>Is there any chance that I could use JFFS2 without any compression?
>>
>>I guess this might me faster for our embedded system!
> 
> 
> Piece of cake.
> Look at fs/jffs2/compr.c, first function. Add a couple of #if 0 and
> you're done.
Thanks!

How about mkfs.jff2? Do I have to change something there as well?

> Do the same with the second function and the kernel image should
> shrink a little.
:o)

Steven

Re: JFFS2 uncompressed?

[Jörn Engel]

On Tue, 8 April 2003 16:33:45 +0200, Steven Scholz wrote:
> 
> How about mkfs.jff2? Do I have to change something there as well?

Yes, you do. Shouldn't be much more complicated, but I didn't do it
myself yet.

Could you post a patch for mkfs.jffs2, when done? Just for the
curious.

Jörn

-- 
Anything that can go wrong, will.
-- Finagle's Law

Re: JFFS2 uncompressed?

[Steven Scholz]

Jörn Engel schrieb:
> On Tue, 8 April 2003 16:33:45 +0200, Steven Scholz wrote:
> 
>>How about mkfs.jff2? Do I have to change something there as well?
> 
> 
> Yes, you do. Shouldn't be much more complicated, but I didn't do it
> myself yet.
> 
> Could you post a patch for mkfs.jffs2, when done? Just for the
> curious.

It's working! :o)

Would you like a define like

#define USE_COMPRESSION

or rather

#define DONT_USE_COMP
or something like that!?

BTW: Where could I put it so mkfs.jffs2 will find it?
Or should we add a command line option to mkfs.jffs2?

Steven

Re: JFFS2 uncompressed?

[Steven Scholz]

[-- Attachment #1: Type: text/plain, Size: 374 bytes --]

Jörn Engel schrieb:
> On Tue, 8 April 2003 16:33:45 +0200, Steven Scholz wrote:
> 
>>How about mkfs.jff2? Do I have to change something there as well?
> 
> 
> Yes, you do. Shouldn't be much more complicated, but I didn't do it
> myself yet.
> 
> Could you post a patch for mkfs.jffs2, when done? Just for the
> curious.

Here you are!

Just a quick hack! :o) Enjoy!

Steven

[-- Attachment #2: jffs2-nocompress.patch --]
[-- Type: text/plain, Size: 2187 bytes --]

Index: util/mkfs.jffs2.c
===================================================================
RCS file: /home/cvs/mtd/util/mkfs.jffs2.c,v
retrieving revision 1.36
diff -p -u -r1.36 mkfs.jffs2.c
--- util/mkfs.jffs2.c	3 Apr 2003 11:49:22 -0000	1.36
+++ util/mkfs.jffs2.c	8 Apr 2003 15:41:15 -0000
@@ -98,6 +98,7 @@ static int squash_uids = 0;
 static int squash_perms = 0;
 static int fake_times = 0;
 static int target_endian = __BYTE_ORDER;
+static int do_compress = 1;
 static const char *const app_name = "mkfs.jffs2";
 static const char *const memory_exhausted = "memory exhausted";
 
@@ -691,6 +692,9 @@ unsigned char jffs2_compress(unsigned ch
 {
 	int ret;
 
+	if (! do_compress)
+		return JFFS2_COMPR_NONE; /* We don't want to compress */
+
 	ret = jffs2_zlib_compress(data_in, cpage_out, datalen, cdatalen);
 	if (!ret) {
 		return JFFS2_COMPR_ZLIB;
@@ -1177,6 +1181,7 @@ static struct option long_options[] = {
 	{"no-cleanmarkers", 0, NULL, 'n'},
 	{"cleanmarker", 1, NULL, 'c'},
 	{"squash", 0, NULL, 'q'},
+	{"no-compression", 0, NULL, 'u'},
 	{"squash-uids", 0, NULL, 'U'},
 	{"squash-perms", 0, NULL, 'P'},
 	{"faketime", 0, NULL, 'f'},
@@ -1202,6 +1207,7 @@ static char *helptext =
 	"  -D, --devtable=FILE    Use the named FILE as a device table file\n"
 	"  -f, --faketime         Change all file times to '0' for regression testing\n"
 	"  -q, --squash           Squash permissions and owners making all files be owned by root\n"
+	"  -u, --no-compression   Don't use any compression\n"
 	"  -U, --squash-uids      Squash owners making all files be owned by root\n"
 	"  -P, --squash-perms     Squash permissions on all files\n"
 	"  -h, --help             Display this help text\n"
@@ -1220,7 +1226,7 @@ int main(int argc, char **argv)
 	struct filesystem_entry *root;
 
 	while ((opt = getopt_long(argc, argv, 
-					"D:d:r:s:o:qUPfh?vVe:lbp::nc:", long_options, &c)) >= 0) 
+					"D:d:r:s:o:quUPfh?vVe:lbp::nc:", long_options, &c)) >= 0) 
 	{
 		switch (opt) {
 			case 'D':
@@ -1256,6 +1262,10 @@ int main(int argc, char **argv)
 			case 'q':
 				squash_uids = 1;
 				squash_perms = 1;
+				break;
+
+			case 'u':
+				do_compress = 0;
 				break;
 
 			case 'U':

Re: JFFS2 uncompressed?

[Steven Scholz]

Hallo Jörn,

hab grad den Patch gesendet. Bekam aber ein

--------------
Your mail to 'linux-mtd' with the subject

     Re: JFFS2 uncompressed?

Is being held until the list moderator can review it for approval.

The reason it is being held:

     Message has a suspicious header

Either the message will get posted to the list, or you will receive
notification of the moderator's decision.  If you would like to cancel
this posting, please visit the following URL:
------------------

Hast Du den Patch bekommen?

Steven

Re: JFFS2 uncompressed?

[Jörn Engel]

On Tue, 8 April 2003 17:06:34 +0200, Steven Scholz wrote:
> 
> It's working! :o)

The good thing about clean code is that it's so easy to add ugly code
on top of it. :)

> Would you like a define like
> 
> #define USE_COMPRESSION
> 
> or rather
> 
> #define DONT_USE_COMP
> or something like that!?

The change is simple enough already, no neef for more #ifdefs.

There is currently ongoing discussion about implementing conditional
compression in several places. Either per compile, as you did it, per
mount point or per inode.

Per inode is the most flexible, was already prepared a long time ago
and is imo the way. If/when this is done, the only reason to use your
simpler approach is to shrink the kernel image.

In short: don't bother.

> BTW: Where could I put it so mkfs.jffs2 will find it?
> Or should we add a command line option to mkfs.jffs2?

I'd prefer a command line option.

Jörn

-- 
The grand essentials of happiness are: something to do, something to
love, and something to hope for.
-- Allan K. Chalmers

Re: JFFS2 uncompressed?

[Jörn Engel]

On Tue, 8 April 2003 17:43:07 +0200, Steven Scholz wrote:
> >
> >Could you post a patch for mkfs.jffs2, when done? Just for the
> >curious.
> 
> Here you are!
> 
> Just a quick hack! :o) Enjoy!

If you call this a quick hack, I'd like to see your clean code.
Thanks!

> Index: util/mkfs.jffs2.c
> ===================================================================
> RCS file: /home/cvs/mtd/util/mkfs.jffs2.c,v
> retrieving revision 1.36
> diff -p -u -r1.36 mkfs.jffs2.c
> --- util/mkfs.jffs2.c	3 Apr 2003 11:49:22 -0000	1.36
> +++ util/mkfs.jffs2.c	8 Apr 2003 15:41:15 -0000
> @@ -98,6 +98,7 @@ static int squash_uids = 0;
>  static int squash_perms = 0;
>  static int fake_times = 0;
>  static int target_endian = __BYTE_ORDER;
> +static int do_compress = 1;
>  static const char *const app_name = "mkfs.jffs2";
>  static const char *const memory_exhausted = "memory exhausted";
>  
> @@ -691,6 +692,9 @@ unsigned char jffs2_compress(unsigned ch
>  {
>  	int ret;
>  
> +	if (! do_compress)
> +		return JFFS2_COMPR_NONE; /* We don't want to compress */
> +
>  	ret = jffs2_zlib_compress(data_in, cpage_out, datalen, cdatalen);
>  	if (!ret) {
>  		return JFFS2_COMPR_ZLIB;
> @@ -1177,6 +1181,7 @@ static struct option long_options[] = {
>  	{"no-cleanmarkers", 0, NULL, 'n'},
>  	{"cleanmarker", 1, NULL, 'c'},
>  	{"squash", 0, NULL, 'q'},
> +	{"no-compression", 0, NULL, 'u'},
>  	{"squash-uids", 0, NULL, 'U'},
>  	{"squash-perms", 0, NULL, 'P'},
>  	{"faketime", 0, NULL, 'f'},
> @@ -1202,6 +1207,7 @@ static char *helptext =
>  	"  -D, --devtable=FILE    Use the named FILE as a device table file\n"
>  	"  -f, --faketime         Change all file times to '0' for regression testing\n"
>  	"  -q, --squash           Squash permissions and owners making all files be owned by root\n"
> +	"  -u, --no-compression   Don't use any compression\n"
>  	"  -U, --squash-uids      Squash owners making all files be owned by root\n"
>  	"  -P, --squash-perms     Squash permissions on all files\n"
>  	"  -h, --help             Display this help text\n"
> @@ -1220,7 +1226,7 @@ int main(int argc, char **argv)
>  	struct filesystem_entry *root;
>  
>  	while ((opt = getopt_long(argc, argv, 
> -					"D:d:r:s:o:qUPfh?vVe:lbp::nc:", long_options, &c)) >= 0) 
> +					"D:d:r:s:o:quUPfh?vVe:lbp::nc:", long_options, &c)) >= 0) 
>  	{
>  		switch (opt) {
>  			case 'D':
> @@ -1256,6 +1262,10 @@ int main(int argc, char **argv)
>  			case 'q':
>  				squash_uids = 1;
>  				squash_perms = 1;
> +				break;
> +
> +			case 'u':
> +				do_compress = 0;
>  				break;
>  
>  			case 'U':

David, I don't see any obvious mistakes. Is it ok to apply?

Jörn

-- 
With a PC, I always felt limited by the software available. On Unix, 
I am limited only by my knowledge.
-- Peter J. Schoenster

Re: JFFS2 uncompressed?

[Steven Scholz]

Jörn Engel schrieb:
> On Tue, 8 April 2003 17:43:07 +0200, Steven Scholz wrote:
> 
>>>Could you post a patch for mkfs.jffs2, when done? Just for the
>>>curious.
>>
>>Here you are!
>>
>>Just a quick hack! :o) Enjoy!
> 
> 
> If you call this a quick hack, I'd like to see your clean code.
> Thanks!

;-) I am flattered! Thanks!

Quick meant I did not try to use "--no-compress" only "-u" ...

But I'd love to see my name in the MTD Hall of Fame though! ;-)

Cheers,

Steven

Re: JFFS2 uncompressed?

[David Woodhouse]

On Tue, 2003-04-08 at 17:05, Steven Scholz wrote:
> But I'd love to see my name in the MTD Hall of Fame though! ;-)

Send me a SSH key and you can commit your patch and at least get your
name in the $Id$ tag and the changelog... :)

-- 
dwmw2

Re: JFFS2 uncompressed?

[Thayne Harbaugh]

[-- Attachment #1: Type: text/plain, Size: 514 bytes --]

On Tue, 2003-04-08 at 10:08, David Woodhouse wrote:
> On Tue, 2003-04-08 at 17:05, Steven Scholz wrote:
> > But I'd love to see my name in the MTD Hall of Fame though! ;-)

For us, it's the cheap thrill of having our name in the MTD Hall of Fame
. . .

> 
> Send me a SSH key and you can commit your patch and at least get your
> name in the $Id$ tag and the changelog... :)

For Dave, it's that he's just too lazy to commit the patches we send him
. . .

;^)

-- 
Thayne Harbaugh
Linux Networx

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 232 bytes --]

Re: JFFS2 uncompressed?

[David Woodhouse]

On Tue, 2003-04-08 at 17:30, Thayne Harbaugh wrote:
> For Dave, it's that he's just too lazy to commit the patches we send him

So true. And if you commit it yourself I know who to blame when three
weeks later it turns out to be broken. :)

-- 
dwmw2

Re: JFFS2 uncompressed?

[Steven Scholz]

David Woodhouse schrieb:
> On Tue, 2003-04-08 at 17:05, Steven Scholz wrote:
> 
>>But I'd love to see my name in the MTD Hall of Fame though! ;-)
> 
> 
> Send me a SSH key and you can commit your patch and at least get your
> name in the $Id$ tag and the changelog... :)

Do you need an entry for the man page as well?

Steven