Re: 2.5.1-pre5 not easy to boot with devfs

[Keith Owens]

On Sun, 2 Dec 2001 11:59:34 -0500 (EST), 
Alexander Viro <[e-mail]> wrote:
>On Sun, 2 Dec 2001, Pierre Rousselet wrote:
>
>> Here is the final (i hope) verdict of my devfs testbox :
>> 
>> 2.4.16 with devfsd-1.3.18/1.3.20 : OK
>> 2.4.17-pre1         "            : Broken
>> 2.5.1-pre1          "            : OK
>> 2.5.1-pre2 with or without v200  : Broken
>> 2.5.1-pre5          "            : Broken
>
>IOW, merge of new devfs code (2.4.17-pre1 in -STABLE, 2.5.1-pre2 in -CURRENT).
>
>We really need CONFIG_DEBUG_* forced if CONFIG_DEVFS_FS is set.  Otherwise
>we'll be getting tons of bug reports due to silent memory corruption.
>
>Keith, is there a decent way to do that?  For 2.4.17 it would help a lot...

Against 2.4.17-pre2, untested.  Revert before 2.4.17.

Index: 17-pre2.1/fs/Config.in
--- 17-pre2.1/fs/Config.in Tue, 13 Nov 2001 08:45:38 +1100 kaos (linux-2.4/m/b/39_Config.in 1.2.1.2.1.7 644)
+++ 17-pre2.1(w)/fs/Config.in Mon, 03 Dec 2001 09:54:58 +1100 kaos (linux-2.4/m/b/39_Config.in 1.2.1.2.1.7 644)
@@ -63,7 +63,10 @@ bool '/proc file system support' CONFIG_
 
 dep_bool '/dev file system support (EXPERIMENTAL)' CONFIG_DEVFS_FS $CONFIG_EXPERIMENTAL
 dep_bool '  Automatically mount at boot' CONFIG_DEVFS_MOUNT $CONFIG_DEVFS_FS
-dep_bool '  Debug devfs' CONFIG_DEVFS_DEBUG $CONFIG_DEVFS_FS
+if [ "$CONFIG_DEVFS_FS" = "y" ] ; then \
+   define_bool CONFIG_DEVFS_DEBUG y
+fi
+# dep_bool '  Debug devfs' CONFIG_DEVFS_DEBUG $CONFIG_DEVFS_FS
 
 # It compiles as a module for testing only.  It should not be used
 # as a module in general.  If we make this "tristate", a bunch of people