XPost: uk.comp.os.linux   
   From: java@evij.com.invalid   
      
   On 2025-06-02 13:21, Theo wrote:   
   >   
   > In uk.comp.os.linux Java Jive  wrote:   
   >>   
   >> RAMDISK: Compressed image found at block 0   
   >>   
   >> # Above is normal   
   >> # Below is crash   
   >>   
   >> EXT3-fs: Magic mismatch, very weird !   
   >> List of all partitions:   
   >> 0800 3907018584 sda driver: sd   
   >>     0801     498688 sda1   
   >>     0802 3906518016 sda2   
   >> 1f00        128 mtdblock0 (driver?)   
   >> 1f01       1792 mtdblock1 (driver?)   
   >> 1f02       1664 mtdblock2 (driver?)   
   >> 1f03        448 mtdblock3 (driver?)   
   >> 1f04         48 mtdblock4 (driver?)   
   >> 1f05          8 mtdblock5 (driver?)   
   >> 1f06          8 mtdblock6 (driver?)   
   >> No filesystem could mount root, tried:  ext3 ext2 vfat fuseblk   
   >> Kernel panic - not syncing: VFS: Unable to mount root fs on   
   >> unknown-block(1,0)   
   >>   
   >> # Below *would* have been a normal continuation for a successful boot   
   >>   
   >> VFS: Mounted root (ext2 filesystem).   
   >   
   > So you appear to be making an ext2 FS and gzipping it.  Do you get the   
   > 'RAMDISK: Compressed image found' in the crash scenario?   
      
   Yes, doing a diff between the successful and kernel panic boots, the   
   RAMDISK line is the last line common between the two, thereafter they   
   diverge as explained.   
      
   > Searching on "Magic mismatch, very weird" comes up with some threads. One is   
   > hardware failure, the other is about using a non-1k blocksize with (old)   
   mke2fs   
   > and a 2007-era ramdisk implementation that doesn't support other than 1k:   
   > https://sourceforge.net/p/e2fsprogs/bugs/175/#b0df   
   >   
   > Perhaps you could try -b1024 on the mkfs.ext2 command?  Or experiment with   
   > other blocksizes?   
      
   Thanks, may try that later this afternoon, but what baffles me is that I   
   think I've completely followed the procedure in the original scripts, so   
   why is the result so different?   
      
   BTW, my attempt to rebuild from scratch is failing also, when building   
   Busybox:   
      
      CC      loginutils/passwd.o   
   loginutils/passwd.c: In function ‘passwd_main’:   
   loginutils/passwd.c:93:16: error: storage size of ‘rlimit_fsize’ isn’t   
   known   
      struct rlimit rlimit_fsize;   
                    ^   
   loginutils/passwd.c:180:2: warning: implicit declaration of function   
   ‘setrlimit’ [-Wimplicit-function-declaration]   
      setrlimit(RLIMIT_FSIZE, &rlimit_fsize);   
      ^   
   loginutils/passwd.c:180:12: error: ‘RLIMIT_FSIZE’ undeclared (first use   
   in this function)   
      setrlimit(RLIMIT_FSIZE, &rlimit_fsize);   
                ^   
   loginutils/passwd.c:180:12: note: each undeclared identifier is reported   
   only once for each function it appears in   
   loginutils/passwd.c:93:16: warning: unused variable ‘rlimit_fsize’   
   [-Wunused-variable]   
      struct rlimit rlimit_fsize;   
                    ^   
   make[2]: *** [loginutils/passwd.o] Error 1   
   make[1]: *** [loginutils] Error 2   
   make[1]: Leaving directory   
   `/home/devel/zyxel/NSA-221-GPL/TestBuild/trunk/sysapps/busybox-1.17.2'   
   make: *** [busybox_initrd] Error 2   
      
   Don't understand it, can't remember ever having that error before in any   
   of the previous builds.   
      
   --   
      
   Fake news kills!   
      
   I may be contacted via the contact address given on my website:   
   www.macfh.co.uk   
      
   --- SoupGate-DOS v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)