[PATCH] drm/radeon/kms: fix suspend on rv530 asics

Dave Airlie airlied at gmail.com
Wed Apr 13 13:16:48 PDT 2011


On Thu, Apr 14, 2011 at 12:52 AM, Alex Deucher <alexdeucher at gmail.com> wrote:
> On Wed, Apr 13, 2011 at 10:46 AM, Jerome Glisse <j.glisse at gmail.com> wrote:
>> On Tue, Apr 12, 2011 at 1:33 PM, Alex Deucher <alexdeucher at gmail.com> wrote:
>>> Apparently only rv515 asics need the workaround
>>> added in f24d86f1a49505cdea56728b853a5d0a3f8e3d11
>>> (drm/radeon/kms: fix resume regression for some r5xx laptops).
>>>
>>> Fixes:
>>> https://bugs.freedesktop.org/show_bug.cgi?id=34709
>>>
>>> Signed-off-by: Alex Deucher <alexdeucher at gmail.com>
>>> Cc: stable at kernel.org
>>> ---
>>>  drivers/gpu/drm/radeon/atom.c |    6 +++++-
>>>  1 files changed, 5 insertions(+), 1 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/radeon/atom.c b/drivers/gpu/drm/radeon/atom.c
>>> index 258fa5e..d71d375 100644
>>> --- a/drivers/gpu/drm/radeon/atom.c
>>> +++ b/drivers/gpu/drm/radeon/atom.c
>>> @@ -32,6 +32,7 @@
>>>  #include "atom.h"
>>>  #include "atom-names.h"
>>>  #include "atom-bits.h"
>>> +#include "radeon.h"
>>>
>>>  #define ATOM_COND_ABOVE                0
>>>  #define ATOM_COND_ABOVEOREQUAL 1
>>> @@ -101,7 +102,9 @@ static void debug_print_spaces(int n)
>>>  static uint32_t atom_iio_execute(struct atom_context *ctx, int base,
>>>                                 uint32_t index, uint32_t data)
>>>  {
>>> +       struct radeon_device *rdev = ctx->card->dev->dev_private;
>>>        uint32_t temp = 0xCDCDCDCD;
>>> +
>>>        while (1)
>>>                switch (CU8(base)) {
>>>                case ATOM_IIO_NOP:
>>> @@ -112,7 +115,8 @@ static uint32_t atom_iio_execute(struct atom_context *ctx, int base,
>>>                        base += 3;
>>>                        break;
>>>                case ATOM_IIO_WRITE:
>>> -                       (void)ctx->card->ioreg_read(ctx->card, CU16(base + 1));
>>> +                       if (rdev->family == CHIP_RV515)
>>> +                               (void)ctx->card->ioreg_read(ctx->card, CU16(base + 1));
>>>                        ctx->card->ioreg_write(ctx->card, CU16(base + 1), temp);
>>>                        base += 3;
>>>                        break;
>>> --
>>> 1.7.1.1
>>>
>>
>>
>> So this patch enable io write only for one family ? This looks utterly strange.
>
> No, it just does a read before write for rv515.  I don't know why it
> needs it, but it seems to.
>

Yeah I really wish I knew why either,

Thinkpad T60 with X1300, no resume without this, it failed in the
memory initialisation table. this was the only thing I could find to
fix it.

My x1300 desktop card works fine without this.

Dave.


More information about the dri-devel mailing list