Searched hist:"1 ffbc51a0d00e52983c70aa7c8dbc7b621d6287d" (Results 1 – 3 of 3) sorted by relevance
| /openbmc/linux/arch/sparc/include/asm/ |
| H A D | leon_amba.h | diff 1ffbc51a0d00e52983c70aa7c8dbc7b621d6287d Mon Jun 10 01:53:28 CDT 2013 Andreas Larsson <andreas@gaisler.com> sparc32, leon: Remove separate "ticker" timer for SMP
This reduces the need from two timers to one timer.
Moreover, without this patch, when the "ticker" timer triggers timer_cs_read via
tick_periodic it reads the value of the usual timer it can get an wrapped timer
value without timer_cs_internal_counter having been updated leading to the clock
going backwards. This effectively hangs one cpu that gets stuck in
update_wall_time with an offset slightly smaller than 0xffffffffffffffff.
Signed-off-by: Andreas Larsson <andreas@gaisler.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| H A D | leon.h | diff 1ffbc51a0d00e52983c70aa7c8dbc7b621d6287d Mon Jun 10 01:53:28 CDT 2013 Andreas Larsson <andreas@gaisler.com> sparc32, leon: Remove separate "ticker" timer for SMP
This reduces the need from two timers to one timer.
Moreover, without this patch, when the "ticker" timer triggers timer_cs_read via
tick_periodic it reads the value of the usual timer it can get an wrapped timer
value without timer_cs_internal_counter having been updated leading to the clock
going backwards. This effectively hangs one cpu that gets stuck in
update_wall_time with an offset slightly smaller than 0xffffffffffffffff.
Signed-off-by: Andreas Larsson <andreas@gaisler.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|
| /openbmc/linux/arch/sparc/kernel/ |
| H A D | leon_kernel.c | diff 1ffbc51a0d00e52983c70aa7c8dbc7b621d6287d Mon Jun 10 01:53:28 CDT 2013 Andreas Larsson <andreas@gaisler.com> sparc32, leon: Remove separate "ticker" timer for SMP
This reduces the need from two timers to one timer.
Moreover, without this patch, when the "ticker" timer triggers timer_cs_read via
tick_periodic it reads the value of the usual timer it can get an wrapped timer
value without timer_cs_internal_counter having been updated leading to the clock
going backwards. This effectively hangs one cpu that gets stuck in
update_wall_time with an offset slightly smaller than 0xffffffffffffffff.
Signed-off-by: Andreas Larsson <andreas@gaisler.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
|