Skip to contentSkip to conversation

config: Fix CONFIG_LFS option

PATCH In review Trond Myklebust 3 participants 4 messages Mar 28, 2006

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 config: Fix CONFIG_LFS option

The help text says that if you select CONFIG_LBD, then it will automatically select CONFIG_LFS. Nope... That isn't currently the case.

- Fix CONFIG_LBD to automatically select CONFIG_LFS
- Get rid of the cruft in the help text mentioning CONFIG_LBD
- Tell unsure users to select CONFIG_LFS.

---

 block/Kconfig |   11 ++++++-----
 1 files changed, 6 insertions(+), 5 deletions(-)
block/Kconfig +6 -5
index 5536839..8b4762c 100644
@@ -6,6 +6,7 @@ #for instance.
 config LBD
 	bool "Support for Large Block Devices"
 	depends on X86 || (MIPS && 32BIT) || PPC32 || (S390 && !64BIT) || SUPERH || UML
+	select LSF
 	help
 	  Say Y here if you want to attach large (bigger than 2TB) discs to
 	  your machine, or if you want to have a raid or loopback device
@@ -27,10 +28,10 @@ config BLK_DEV_IO_TRACE
 config LSF
 	bool "Support for Large Single Files"
 	depends on X86 || (MIPS && 32BIT) || PPC32 || ARCH_S390_31 || SUPERH || UML
-	default n
-	help
-	  When CONFIG_LBD is disabled, say Y here if you want to
-	  handle large file(bigger than 2TB), otherwise say N.
-	  When CONFIG_LBD is enabled, Y is set automatically.
+	help
+	  Say Y here if you want to be able to handle very large files (bigger
+	  than 2TB), otherwise say N.
+
+	  If unsure, say Y.
 
 source block/Kconfig.iosched
#1 lore raw
Quoted 4 lines · Trond Myklebust

The help text says that if you select CONFIG_LBD, then it will automatically select CONFIG_LFS. Nope... That isn't currently the case.

I'm not sure your patch makes anything much better, though.

Why does CONFIG_LSF exist in the first place? Afaik, it only affects a totally not-very-interesting thing (blkcnt_t) for a totally not very interesting feature (the number of people who want single files >2TB is likely not very big).

Having it auto-selected by LBD sounds insane, since LBD is likely more interesting than LSF itsef is. It would make more sense to go the other way (have LSF auto-select LBD).

Linus

#2 lore raw
Quoted 9 lines · Linus Torvalds
Quoted 4 lines · Trond Myklebust

The help text says that if you select CONFIG_LBD, then it will automatically select CONFIG_LFS. Nope... That isn't currently the case.

Quoted 6 lines

I'm not sure your patch makes anything much better, though.

Why does CONFIG_LSF exist in the first place? Afaik, it only affects a totally not-very-interesting thing (blkcnt_t) for a totally not very interesting feature (the number of people who want single files >2TB is likely not very big).

Yes, it's strictly a bugfix, but it's not a very big bug. Perhaps we should turn it on all the time and not have the config option, but it does have some cost in text size, data size and instruction count.

Quoted 3 lines

Having it auto-selected by LBD sounds insane, since LBD is likely more interesting than LSF itsef is. It would make more sense to go the other way (have LSF auto-select LBD).

Spose so. I wonder what distributors will choose to do.

It's a bit odd to have a config option to select whether or not to have a buggy kernel.

#3 lore raw
In reply to Linus Torvalds
Quoted 9 lines · Linus Torvalds
Quoted 4 lines · Trond Myklebust

The help text says that if you select CONFIG_LBD, then it will automatically select CONFIG_LFS. Nope... That isn't currently the case.

I'm not sure your patch makes anything much better, though.

Quoted 8 lines

Why does CONFIG_LSF exist in the first place? Afaik, it only affects a totally not-very-interesting thing (blkcnt_t) for a totally not very interesting feature (the number of people who want single files >2TB is likely not very big).

Having it auto-selected by LBD sounds insane, since LBD is likely more interesting than LSF itsef is. It would make more sense to go the other way (have LSF auto-select LBD).

NFSv3 and CIFS are two examples of commonly used filesystems that don't care a hoot for CONFIG_LBD, but that still want to be able to support large values for inode->i_blocks.

Cheers,
Trond