https://bugzilla.novell.com/show_bug.cgi?id=672450
https://bugzilla.novell.com/show_bug.cgi?id=672450#c9
--- Comment #9 from Jiri Slaby
Jiri? You may remove the files
/lib/mkinitrd/scripts/boot-blogd.sh /lib/mkinitrd/scripts/boot-killblogd.sh /lib/mkinitrd/scripts/boot-killblogd2.sh
and change within
/lib/mkinitrd/scripts/boot-start.s
the line
exec < /dev/console > /dev/console 2>&1
to
exec < /dev/console > /dev/kmsg 2>&1
Yes, this stops blogd to flood the console.
... Nevertheless the serial driver in the kernel should not fool the usage of the system call write(2).
Nope, it's blogd who cripples. With this change (write(2) interception in the kernel): diff --git a/fs/read_write.c b/fs/read_write.c index 5520f8a..ecae48a 100644 --- a/fs/read_write.c +++ b/fs/read_write.c @@ -384,6 +384,14 @@ ssize_t vfs_write(struct file *file, const char __user *buf, size_t count, loff_ inc_syscw(current); } + if (!strcmp(current->comm, "blogd")) { + int a; + printk("%s: cnt=%lu ret=%ld str=", __func__, count, ret); + for (a = 0; a < min(count, 16LU); a++) + printk(KERN_CONT "%c", buf[a] == '\n' ? 'X' : buf[a]); + printk(KERN_CONT "\n"); + } + return ret; } The output looks like (i.e. blogd write the same messages several times even though the retval equals to what it wanted kernel to write...): [ 2.612154] vfs_write: cnt=78 ret=78 str=^MBoot logging st [ 2.616610] vfs_write: cnt=78 ret=78 str=^MBoot logging st [ 2.621521] vfs_write: cnt=5 ret=5 str=1095X [ 2.634171] vfs_write: cnt=118 ret=118 str=^MBoot logging st [ 2.640975] vfs_write: cnt=118 ret=118 str=^MBoot logging st [ 2.646736] vfs_write: cnt=118 ret=118 str=^MBoot logging st -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.