Messages in this thread Patch in this message | | | Date | Fri, 4 Feb 2000 15:02:06 +0100 (CET) | From | Ingo Molnar <> | Subject | [patch] keyboard-2.3.42-A0 |
| |
2.3.41 broke my keyboard, the reason is the less forgiving nature of dealing with keyboard errors. The attached patch makes send_data() a bit more patient, until the reason for the errors is found.
it seems a couple of keyboard commands at bootup fail on my box - this magically gets fixed later on. (something times out? is my box too fast?)
prior to 2.3.41 i was getting such timeouts:
Keyboard timeout[2] Keyboard timeout[2]
two very different systems show this failure mode. Both have keyboards connected over the PS2 port. One is a true PS2 keyboard, the other one is an old-style one converted with a PS2 converter cable. Basically the only similarity in those systems is that both are SMP systems, use the PS2 port and are rather fast. (and run similar kernel .config-s.) No USB enabled in the kernel.
i happen to have logs which shows 2.3.30 showing the same boot-time keyboard timeout. Nevertheless i have another 2.3.30 log which shows no such timeout. Unfortunately i was working on many patches at once so it's not clear what kind of 2.3.30 kernel it was. Neither do i have all historic logs so it's not clear wether prior kernels ever have shown this problem. (but 2.3.30-ish appears to be the first occasion from the logs i have)
let me know if there is any patch or other debugging suggestion i should try.
-- mingo --- linux/drivers/char/pc_keyb.c.orig2 Fri Feb 4 05:38:13 2000 +++ linux/drivers/char/pc_keyb.c Fri Feb 4 05:39:15 2000 @@ -505,12 +505,7 @@ #ifdef KBD_REPORT_TIMEOUTS printk(KERN_WARNING "keyboard: Timeout - AT keyboard not present?\n"); #endif - /* - * Investigate why some keyboards fail for - * one or two commands but then magically - * recover. - */ -// kbd_exists = 0; + kbd_exists = 1; return 0; } }
| |