Syslinux - Aug 2005 - SYSLINUX 3.11-pre5 -- release candidate -- release on Tuesday

SYSLINUX 3.11-pre5 is hereby a bug fix release candidate.  If I don't 
receive any problem reports, I will release it as-is on Tuesday, August 30.

http://www.kernel.org/pub/linux/utils/boot/syslinux/Testing/

	-hpa

H. Peter Anvin wrote:
> SYSLINUX 3.11-pre5 is hereby a bug fix release candidate.  If I don't 
> receive any problem reports, I will release it as-is on Tuesday, 
> August 30.
>
> http://www.kernel.org/pub/linux/utils/boot/syslinux/Testing/
>
>     -hpa
>
Im seeing a truncation of the boot-line at 256, despite the 511 
command_line_len in the code,
and my pxelinux.cfg/default, which contains:

LABEL   H 2.6.13-rc6-mm2-kgdb-v1
  MENU LABEL    ^h.  2.6.13-rc6-mm2-kgdb-v1
  KERNEL        vmlinuz-2.6.13-rc6-mm2-kgdb-v1
  APPEND        console=ttyS0,115200n81 root=/dev/nfs 
nfsroot=192.168.42.1:/nfshost/truck 
nfsaddrs=192.168.42.100:192.168.42.1:192.168.42.1:255.255.255.0:soekris:eth0 
panic=5 initrd=initrd-2.6.13-rc6-mm2-kgdb-v1.img 
kgdboe=7000 at 192.168.42.100/, at 192.168.42.1/


on-console, it shows:

                          Press [Tab] to edit options

 > vmlinuz-2.6.13-rc6-mm2-kgdb-v1 console=ttyS0,115200n81 root=/dev/nfs 
nfsroot=192.168.42.1:/nfshost/truck 
nfsaddrs=192.168.42.100:192.168.42.1:192.168.42.1:255.255.255.0:soekris:eth0 
panic=5 initrd=initrd-2.6.13-rc6-mm2-kgdb-v1.img kgdboe=7000 at 192.168.42.1

note the lack of '/, at 192.168.42.1/'  
It almost looks like a parsing error, but it is a 256 truncation thing.
as shown by adding a nonsense boot param 'booble'

 > vmlinuz-2.6.13-rc6-mm2-kgdb-v1 console=ttyS0,115200n81 root=/dev/nfs 
nfsroot=192.168.42.1:/nfshost/truck 
nfsaddrs=192.168.42.100:192.168.42.1:192.168.42.1:255.255.255.0:soekris:eth0 
panic=5 initrd=initrd-2.6.13-rc6-mm2-kgdb-v1.img booble kgdboe=7000 at 192.1



Ill start working backwards to find the version responsible.

Jim Cromie wrote:
>
>
> on-console, it shows:
>
turns out is a display problem (pxelinux - I dont use syslinux)
- the kernel boot-line sees it (Ive patched my kernel to 511)
I suspect this has never worked, as its never been needed with current 
kernel limit.

BTW - you said a while back that larger command_line_boot-lines were
causing some problems somewhere.  Is that still the case ?

Ive been using this, no problems so far.
Signed-off-by:  Jim Cromie <jim.cromie at gmail.com>

--- lnew/include/asm-i386/setup.h~      2005-06-07 13:34:14.000000000 -0600
+++ lnew/include/asm-i386/setup.h       2005-06-08 13:33:12.000000000 -0600
@@ -17,7 +17,7 @@
 #define MAX_NONPAE_PFN (1 << 20)

 #define PARAM_SIZE 4096
-#define COMMAND_LINE_SIZE 256
+#define COMMAND_LINE_SIZE 512

 #define OLD_CL_MAGIC_ADDR      0x90020
 #define OLD_CL_MAGIC           0xA33F
--- lnew/include/asm-i386/param.h~      2005-06-07 13:35:23.000000000 -0600
+++ lnew/include/asm-i386/param.h       2005-06-08 13:33:29.000000000 -0600
@@ -20,6 +20,6 @@
 #endif

 #define MAXHOSTNAMELEN 64      /* max length of hostname */
-#define COMMAND_LINE_SIZE 256
+#define COMMAND_LINE_SIZE 512

 #endif

Jim Cromie wrote:
> Jim Cromie wrote:
>
>>
>>
>> on-console, it shows:
>>
>
> turns out is a display problem (pxelinux - I dont use syslinux)
> - the kernel boot-line sees it (Ive patched my kernel to 511)
> I suspect this has never worked, as its never been needed with current 
> kernel limit.
>
and heres a fix.  -p0 patch to prevent line-wrap
Apologies, shoulda all been one email.  { it will probly happen again :-}



diff -u com32/modules/menu.h~ com32/modules/menu.h
--- com32/modules/menu.h~       2005-08-25 18:51:49.000000000 -0600
+++ com32/modules/menu.h        2005-08-26 22:47:01.000000000 -0600
@@ -28,7 +28,7 @@
   unsigned char hotkey;
 };

-#define MAX_CMDLINE_LEN         256
+#define MAX_CMDLINE_LEN         512

 #define MAX_ENTRIES    4096    /* Oughta be enough for anybody */
 extern struct menu_entry menu_entries[];