| Message ID | 1310370069-20901-1-git-send-email-l.majewski@samsung.com |
|---|---|
| State | Changes Requested |
| Delegated to: | Minkyu Kang |
| Headers | show |
Hi Lukasz, Le 11/07/2011 09:41, Lukasz Majewski a écrit : > Signed-off-by: Lukasz Majewski<l.majewski@samsung.com> > Cc: Minkyu Kang<mk7.kang@samsung.com> > Cc: Aneesh V<aneesh@ti.com> > --- Can you please submit a V2 patch in which the commit message provides a rationale for turning the data cache off? People who want to turn it on later will benefit from the info. Amicalement,
On Thu, 14 Jul 2011 16:00:23 +0200 Albert ARIBAUD <albert.u.boot@aribaud.net> wrote: > Hi Lukasz, > > Le 11/07/2011 09:41, Lukasz Majewski a écrit : > > Signed-off-by: Lukasz Majewski<l.majewski@samsung.com> > > Cc: Minkyu Kang<mk7.kang@samsung.com> > > Cc: Aneesh V<aneesh@ti.com> > > --- > > Can you please submit a V2 patch in which the commit message provides > a rationale for turning the data cache off? People who want to turn > it on later will benefit from the info. > > Amicalement, Hi Albert, Rationale for turning OFF D cache: 1. Before change introduced in commit SHA1:c2dd0d45540397704de9b13287417d21049d34c6 the D-cache for GONI/Universal_c210 targets was disabled. This change was supposed to preserve this state. 2. I've done some investigation about enabling D-cache support. It turned out, that some drivers (which are already posted or are in the queue for posting to u-boot list) have had problems with D-cache coherency. This commit shall be treated as a temporal solution. After solving cache management (and coherency) problems the D-cache would be definitely enabled.
Hi Lukasz, Le 15/07/2011 09:34, Lukasz Majewski a écrit : > On Thu, 14 Jul 2011 16:00:23 +0200 > Albert ARIBAUD<albert.u.boot@aribaud.net> wrote: > >> Hi Lukasz, >> >> Le 11/07/2011 09:41, Lukasz Majewski a écrit : >>> Signed-off-by: Lukasz Majewski<l.majewski@samsung.com> >>> Cc: Minkyu Kang<mk7.kang@samsung.com> >>> Cc: Aneesh V<aneesh@ti.com> >>> --- >> >> Can you please submit a V2 patch in which the commit message provides >> a rationale for turning the data cache off? People who want to turn >> it on later will benefit from the info. >> >> Amicalement, > > Hi Albert, > > Rationale for turning OFF D cache: > > 1. Before change introduced in commit > SHA1:c2dd0d45540397704de9b13287417d21049d34c6 the D-cache for > GONI/Universal_c210 targets was disabled. This change was supposed to > preserve this state. > > 2. I've done some investigation about enabling D-cache support. It > turned out, that some drivers (which are already posted or are in the > queue for posting to u-boot list) have had problems with D-cache > coherency. > > This commit shall be treated as a temporal solution. After solving > cache management (and coherency) problems the D-cache would be > definitely enabled. Thanks. Except for a typo (Temporal-> Temporary), please put this rationale in the commit message for V2. Amicalement,
diff --git a/include/configs/s5p_goni.h b/include/configs/s5p_goni.h index d648ce8..010428b 100644 --- a/include/configs/s5p_goni.h +++ b/include/configs/s5p_goni.h @@ -74,6 +74,9 @@ /* It should define before config_cmd_default.h */ #define CONFIG_SYS_NO_FLASH 1 +/* Cache D configuration */ +#define CONFIG_SYS_DCACHE_OFF + /* Command definition */ #include <config_cmd_default.h>
Signed-off-by: Lukasz Majewski <l.majewski@samsung.com> Cc: Minkyu Kang <mk7.kang@samsung.com> Cc: Aneesh V <aneesh@ti.com> --- include/configs/s5p_goni.h | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-)