Re: fr1-2.16 patch for 2.6.10 kernel

In article <[email protected]> you wrote:
> sda1 (read) [events: 0000022a]
> md: old uuid 5dc1267b 788f7499 cd49d3c0 d06f0b96
> md: new uuid 5dc1267b 788f7499 cd49d3c0 d06f0b96
> md0: new disk 800001 too old for repair (disk 554 < bitmap 556)

Do you think I ought to change the comparison from "-1" to "-2"? You
seem to get a bitmap that is started 2 events after the disk is
faulted, even though it is started "in the act of faulting".  I suppose
something might sleep and allow that behaviour, or it might be
preemption, or genuine SMP. 

I could simply stamp the time and compare that.


YOu can see where it is in the patch:

....
+                if (disk_events == bitmap_events - 1) {
+                        printk(KERN_WARNING
+                        "%s: warning - new disk %x nearly too old for repair (d
isk %Ld < bitmap %Ld)\n",
+                        mdname(mddev), dev, disk_events, bitmap_events);
+                }
+                if (disk_events < bitmap_events - 1) {
+                        /* new disk is too old! */
...

You might want to try the obvious changes (- 1 -> - 2, twice).

If it's a big deal, I could go for a timestamp.

Peter