Skip to contentSkip to conversation

Linux 2.4.20-pre2 compile error

DISCUSSION Chad Young 6 participants 8 messages Aug 13, 2002

lore
Share & tools
lore

Chad Young

author
#0 lore raw

any idea what causes these errors?

make[3]: Entering directory `/home/skidley/kernel/linux-2.4.20-pre2/fs/partitions'
gcc -D__KERNEL__ -I/home/skidley/kernel/linux-2.4.20-pre2/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=i686 -nostdinc -I /usr/lib/gcc-lib/i386-linux/2.95.4/include -DKBUILD_BASENAME=check -DEXPORT_SYMTAB -c check.c
check.c: In function `devfs_register_disc':
check.c:328: structure has no member named number' check.c:329: structure has no member named number'
check.c: In function `devfs_register_partitions':
check.c:361: structure has no member named `number'
make[3]: *** [check.o] Error 1
make[3]: Leaving directory `/home/skidley/kernel/linux-2.4.20-pre2/fs/partitions' make[2]: *** [first_rule] Error 2
make[2]: Leaving directory `/home/skidley/kernel/linux-2.4.20-pre2/fs/partitions' make[1]: *** [_subdir_partitions] Error 2
make[1]: Leaving directory `/home/skidley/kernel/linux-2.4.20-pre2/fs'
make: *** [_dir_fs] Error 2

Signature · 4 lines
"I mean they are gonna kill ya so like if ya give em a quick, short, sharp, 
shock they won't do it again. Dig it! I mean he got off lightly cuz I would 
have given him a thrashing. I only hit him once. It was only a difference of 
opinion but really... I mean good manners don't cost nothin do they. Eh?"
#1 lore raw
Quoted 9 lines · Chad Young

any idea what causes these errors?

make[3]: Entering directory `/home/skidley/kernel/linux-2.4.20-pre2/fs/partitions'
gcc -D__KERNEL__ -I/home/skidley/kernel/linux-2.4.20-pre2/include -Wall

(quoted diff) +0 -4
-Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common
-fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=i686
-nostdinc -I /usr/lib/gcc-lib/i386-linux/2.95.4/include
-DKBUILD_BASENAME=check  -DEXPORT_SYMTAB -c check.c
Quoted 6 lines

check.c: In function `devfs_register_disc':
check.c:328: structure has no member named number' check.c:329: structure has no member named number' check.c: In function `devfs_register_partitions':
check.c:361: structure has no member named `number'
...

The following patch made by Christoph Hellwig fixes it:

include/linux/genhd.h +3 -1
@@ -62,7 +62,9 @@ struct hd_struct {
 	unsigned long start_sect;
 	unsigned long nr_sects;
 	devfs_handle_t de;              /* primary (master) devfs entry  */
-
+#ifdef CONFIG_DEVFS_FS
+	int number;
+#endif /* CONFIG_DEVFS_FS */
 #ifdef CONFIG_BLK_STATS
 	/* Performance stats: */
 	unsigned int ios_in_flight;

cu
Adrian

Signature · 3 lines
You only think this is a free country. Like the US the UK spends a lot of
time explaining its a free country because its a police state.
								Alan Cox
#2 lore raw

are you sure about the patch ?

best regards

Quoted 50 lines · Adrian Bunk
Quoted 9 lines · Chad Young

any idea what causes these errors?

make[3]: Entering directory `/home/skidley/kernel/linux-2.4.20-pre2/fs/partitions'
gcc -D__KERNEL__ -I/home/skidley/kernel/linux-2.4.20-pre2/include -Wall

(quoted diff) +0 -4
-Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common
-fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=i686
-nostdinc -I /usr/lib/gcc-lib/i386-linux/2.95.4/include
-DKBUILD_BASENAME=check  -DEXPORT_SYMTAB -c check.c
Quoted 6 lines

check.c: In function `devfs_register_disc':
check.c:328: structure has no member named number' check.c:329: structure has no member named number' check.c: In function `devfs_register_partitions':
check.c:361: structure has no member named `number'
...

The following patch made by Christoph Hellwig fixes it:

include/linux/genhd.h +3 -1
@@ -62,7 +62,9 @@ struct hd_struct {
 	unsigned long start_sect;
 	unsigned long nr_sects;
 	devfs_handle_t de;              /* primary (master) devfs entry  */
-
+#ifdef CONFIG_DEVFS_FS
+	int number;
+#endif /* CONFIG_DEVFS_FS */
 #ifdef CONFIG_BLK_STATS
 	/* Performance stats: */
 	unsigned int ios_in_flight;

cu
Adrian

Signature · 9 lines
You only think this is a free country. Like the US the UK spends a lot of
time explaining its a free country because its a police state.
								Alan Cox

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/
Signature · 2 lines
Stephane Wirtel <[email protected]>
Web : www.linux-mons.be	 "Linux Is Not UniX !!!"
#3 lore raw
Quoted 1 line · Stephane Wirtel

are you sure about the patch ?

It seems to be correct and I can verify both the compile error and that this patch fixes it. What do you consider to be wrong?

Quoted 1 line

best regards

cu
Adrian

Quoted 36 lines · Adrian Bunk
Quoted 35 lines · Chad Young
Quoted 10 lines

any idea what causes these errors?

make[3]: Entering directory `/home/skidley/kernel/linux-2.4.20-pre2/fs/partitions'
gcc -D__KERNEL__ -I/home/skidley/kernel/linux-2.4.20-pre2/include -Wall

(quoted diff) +0 -4
-Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common
-fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=i686
-nostdinc -I /usr/lib/gcc-lib/i386-linux/2.95.4/include
-DKBUILD_BASENAME=check  -DEXPORT_SYMTAB -c check.c
Quoted 6 lines

check.c: In function `devfs_register_disc':
check.c:328: structure has no member named number' check.c:329: structure has no member named number' check.c: In function `devfs_register_partitions':
check.c:361: structure has no member named `number'
...

The following patch made by Christoph Hellwig fixes it:

include/linux/genhd.h +3 -1
@@ -62,7 +62,9 @@ struct hd_struct {
 	unsigned long start_sect;
 	unsigned long nr_sects;
 	devfs_handle_t de;              /* primary (master) devfs entry  */
-
+#ifdef CONFIG_DEVFS_FS
+	int number;
+#endif /* CONFIG_DEVFS_FS */
 #ifdef CONFIG_BLK_STATS
 	/* Performance stats: */
 	unsigned int ios_in_flight;
#4 lore raw

are you sure that the "int number" is employed only by devfs?

Quoted 55 lines · Adrian Bunk
Quoted 1 line · Stephane Wirtel

are you sure about the patch ?

It seems to be correct and I can verify both the compile error and that this patch fixes it. What do you consider to be wrong?

Quoted 1 line

best regards

cu
Adrian

Quoted 22 lines · Adrian Bunk
Quoted 21 lines · Chad Young
Quoted 10 lines

any idea what causes these errors?

make[3]: Entering directory `/home/skidley/kernel/linux-2.4.20-pre2/fs/partitions'
gcc -D__KERNEL__ -I/home/skidley/kernel/linux-2.4.20-pre2/include -Wall

(quoted diff) +0 -4
-Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common
-fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=i686
-nostdinc -I /usr/lib/gcc-lib/i386-linux/2.95.4/include
-DKBUILD_BASENAME=check  -DEXPORT_SYMTAB -c check.c
Quoted 6 lines

check.c: In function `devfs_register_disc':
check.c:328: structure has no member named number' check.c:329: structure has no member named number' check.c: In function `devfs_register_partitions':
check.c:361: structure has no member named `number'
...

The following patch made by Christoph Hellwig fixes it:

Quoted 14 lines
Quoted 4 lines

include/linux/genhd.h struct hd_struct { · line 70

Quoted 10 lines
include/linux/genhd.h +3 -1
@@ -62,7 +62,9 @@ struct hd_struct {
 	unsigned long start_sect;
 	unsigned long nr_sects;
 	devfs_handle_t de;              /* primary (master) devfs entry  */
-
+#ifdef CONFIG_DEVFS_FS
+	int number;
+#endif /* CONFIG_DEVFS_FS */
 #ifdef CONFIG_BLK_STATS
 	/* Performance stats: */
 	unsigned int ios_in_flight;

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Signature · 2 lines
Stephane Wirtel <[email protected]>
Web : www.linux-mons.be	 "Linux Is Not UniX !!!"
#5 lore raw
Quoted 1 line · Stephane Wirtel

are you sure that the "int number" is employed only by devfs?

Yes, it is. And I'm pissed that it neither was depend on the devfs config option
nor had a devfsßrelated named for a long time. Richard just bloats the whole kernel woth devfs crap all over the place.

#6 lore raw
Quoted 4 lines · Christoph Hellwig

On Tue, Aug 13, 2002 at 01:01:38PM +0200, Stephane Wirtel wrote:
Yes, it is. And I'm pissed that it neither was depend on the devfs config option
nor had a devfsßrelated named for a long time. Richard just bloats the whole kernel woth devfs crap all over the place.

I would like to defend devfs a bit, although I know nothing of this particular case.

I have (for ever) HATED the tendency of Unix systems to have one huge /dev directory with, often, thousands of completely useless devnodes in it, many of then with names that are obscure to say the least, and no idea how to find the right one.

For example, recently I wanted to find the combined mouse device on an RH3 system that I haven't got around to switching to devfs; after a lot of searching, including in the kernel sources, I gave up. On defvs it's obvious:
you have /dev/mouse/, and under that "mice", which, being the plural of "mouse", is pretty clear, IMO.

At least devfs does address this, and IMO quite well. I don't at the moment understand the problems people have with it (and if possible I would appreciate a _sensible- discussion), but if anyone wants to return to the 'bad old days' I'm very much against it.

Rant over,

Regards,

Ruth

Signature · 2 lines
Ruth Ivimey-Cook
Software engineer and technical writer.
#7 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wed, 14 Aug 2002 02:21, Ruth Ivimey-Cook wrote:
<snip>

Quoted 5 lines

For example, recently I wanted to find the combined mouse device on an RH3 system that I haven't got around to switching to devfs; after a lot of searching, including in the kernel sources, I gave up. On defvs it's
obvious: you have /dev/mouse/, and under that "mice", which, being the plural of "mouse", is pretty clear, IMO.

Note that this is partly an experience (maybe even cultural) issue.

Given the same problem, I'd think "ah, mouse. Now that is an input device, so I'd expect it to be in /devin/input. Hmm - lots of mouse0, mouse1, etc type entries, and a "mice" entry. That'll probably be the one..."

I personally wouldn't think to look in /dev/mouse. And remember that we speak (mostly) the same language. Maus?

I guess it is just a matter of what you know.

Brad
- --
http://conf.linux.org.au. 22-25Jan2003. Perth, Australia. Birds in Black. -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE9WYKPW6pHgIdAuOMRAlGyAJ9Oh18IJwOG16GhwqO/5Jl/sMRwcwCdFR53 ecU1Ko3Y/s6TOBSuo7aXmGg=
=URqj
-----END PGP SIGNATURE-----