Skip to contentSkip to conversation

Is MI_FLUSH_ENABLE bit 12?

DISCUSSION Keith Packard 5 participants 9 messages Nov 29, 2011 – Dec 6, 2011

lore
Share & tools
lore
#0 lore raw

Just reading through vol1c.4 of the bspec this evening and found something odd.

Bit 11 of MI_MODE is "Invalidate UHPTR enable".
Bit 12 of MI_MODE is "MI_FLUSH Enable"

And, yet, in i915_reg.h:

#define MI_MODE 0x0209c

# define VS_TIMER_DISPATCH				(1 << 6)
# define MI_FLUSH_ENABLE				(1 << 11)

Are we off-by-one on MI_FLUSH_ENABLE? Seems like this would cause serious problems...

Signature · 1 line
[email protected]

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#1 lore raw

On Mon, 28 Nov 2011 18:48:04 -0800, "Keith Packard" <[email protected]> wrote:
Non-text part: multipart/mixed
Non-text part: multipart/signed

Quoted 6 lines

Just reading through vol1c.4 of the bspec this evening and found something odd.

Bit 11 of MI_MODE is "Invalidate UHPTR enable".
Bit 12 of MI_MODE is "MI_FLUSH Enable"

Quoted 8 lines

And, yet, in i915_reg.h:

#define MI_MODE 0x0209c

# define VS_TIMER_DISPATCH				(1 << 6)
# define MI_FLUSH_ENABLE				(1 << 11)

Are we off-by-one on MI_FLUSH_ENABLE? Seems like this would cause serious problems...

I think we are. On the other hand, based on actual behavior plus reading of simulator, I believe that the bit does nothing, regardless.

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#2 lore raw
Quoted 15 lines · Eric Anholt

On Mon, 28 Nov 2011 18:48:04 -0800, "Keith Packard" <[email protected]> wrote:
Non-text part: multipart/mixed
Non-text part: multipart/signed

Quoted 12 lines

Just reading through vol1c.4 of the bspec this evening and found something odd.

Bit 11 of MI_MODE is "Invalidate UHPTR enable".
Bit 12 of MI_MODE is "MI_FLUSH Enable"

And, yet, in i915_reg.h:

#define MI_MODE 0x0209c

# define VS_TIMER_DISPATCH				(1 << 6)
# define MI_FLUSH_ENABLE				(1 << 11)
Quoted 5 lines
Quoted 2 lines

Are we off-by-one on MI_FLUSH_ENABLE? Seems like this would cause serious problems...

I think we are. On the other hand, based on actual behavior plus reading of simulator, I believe that the bit does nothing, regardless.

I do not think so. We've (Chris, I, and perhaps Jesse?) been through this excercise at least twice before, and both times resulted in hangs when we switched to bit 12 on Ironlake, not sure about other platforms.

Ben

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#3 lore raw
Quoted 4 lines · Ben Widawsky

I do not think so. We've (Chris, I, and perhaps Jesse?) been through this excercise at least twice before, and both times resulted in hangs when we switched to bit 12 on Ironlake, not sure about other platforms.

Then, perhaps, a comment in the code would let other people learn from your mis-adventures?

Signature · 1 line
[email protected]

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#4 lore raw
In reply to Ben Widawsky

On Wed, 30 Nov 2011 03:42:00 +0000, Ben Widawsky <[email protected]> wrote:
Non-text part: multipart/signed

Quoted 19 lines · Eric Anholt
Quoted 18 lines · Keith Packard

Non-text part: multipart/mixed
Non-text part: multipart/signed

Quoted 14 lines

Just reading through vol1c.4 of the bspec this evening and found something odd.

Bit 11 of MI_MODE is "Invalidate UHPTR enable".
Bit 12 of MI_MODE is "MI_FLUSH Enable"

And, yet, in i915_reg.h:

#define MI_MODE 0x0209c

# define VS_TIMER_DISPATCH				(1 << 6)
# define MI_FLUSH_ENABLE				(1 << 11)

Are we off-by-one on MI_FLUSH_ENABLE? Seems like this would cause serious problems...

Quoted 6 lines
Quoted 2 lines

I think we are. On the other hand, based on actual behavior plus reading of simulator, I believe that the bit does nothing, regardless.

I do not think so. We've (Chris, I, and perhaps Jesse?) been through this excercise at least twice before, and both times resulted in hangs when we switched to bit 12 on Ironlake, not sure about other platforms.

There is no MI_FLUSH enable bit on Ironlake in my copy of the docs. Bit 12 is MBZ on that hardware.

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#5 lore raw
Quoted 24 lines · On Wed, 30 Nov 2011 13:54:47 -0800 Eric Anholt <[email protected]> wrote:

On Wed, 30 Nov 2011 03:42:00 +0000, Ben Widawsky <[email protected]> wrote:
Non-text part: multipart/signed

Quoted 22 lines · Eric Anholt
Quoted 20 lines · Keith Packard

Non-text part: multipart/mixed
Non-text part: multipart/signed

Quoted 6 lines

Just reading through vol1c.4 of the bspec this evening and found something odd.

Bit 11 of MI_MODE is "Invalidate UHPTR enable".
Bit 12 of MI_MODE is "MI_FLUSH Enable"

Quoted 8 lines

And, yet, in i915_reg.h:

#define MI_MODE 0x0209c

# define VS_TIMER_DISPATCH				(1 << 6)
# define MI_FLUSH_ENABLE				(1 << 11)

Are we off-by-one on MI_FLUSH_ENABLE? Seems like this would cause serious problems...

I think we are. On the other hand, based on actual behavior plus reading of simulator, I believe that the bit does nothing, regardless.

Quoted 6 lines
Quoted 3 lines

I do not think so. We've (Chris, I, and perhaps Jesse?) been through this excercise at least twice before, and both times resulted in hangs when we switched to bit 12 on Ironlake, not sure about other platforms.

There is no MI_FLUSH enable bit on Ironlake in my copy of the docs. Bit 12 is MBZ on that hardware.

It's one of those "been there forever" bits. A comment wouldn't hurt, but we should also put in a doc change request (in fact I think I did that but it went into a black hole).

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#6 lore raw
Quoted 31 lines · Jesse Barnes
Quoted 28 lines · On Wed, 30 Nov 2011 13:54:47 -0800 Eric Anholt <[email protected]> wrote:

On Wed, 30 Nov 2011 03:42:00 +0000, Ben Widawsky <[email protected]> wrote:
Non-text part: multipart/signed

Quoted 25 lines · Eric Anholt
Quoted 15 lines · Keith Packard

Non-text part: multipart/mixed
Non-text part: multipart/signed

Quoted 12 lines

Just reading through vol1c.4 of the bspec this evening and found something odd.

Bit 11 of MI_MODE is "Invalidate UHPTR enable".
Bit 12 of MI_MODE is "MI_FLUSH Enable"

And, yet, in i915_reg.h:

#define MI_MODE 0x0209c

# define VS_TIMER_DISPATCH				(1 << 6)
# define MI_FLUSH_ENABLE				(1 << 11)
Quoted 5 lines
Quoted 2 lines

Are we off-by-one on MI_FLUSH_ENABLE? Seems like this would cause serious problems...

I think we are. On the other hand, based on actual behavior plus reading of simulator, I believe that the bit does nothing, regardless.

I do not think so. We've (Chris, I, and perhaps Jesse?) been through this excercise at least twice before, and both times resulted in hangs when we switched to bit 12 on Ironlake, not sure about other platforms.

Quoted 6 lines
Quoted 2 lines

There is no MI_FLUSH enable bit on Ironlake in my copy of the docs. Bit 12 is MBZ on that hardware.

It's one of those "been there forever" bits. A comment wouldn't hurt, but we should also put in a doc change request (in fact I think I did that but it went into a black hole).

Well, by "been there forever", you mean "introduced in:

commit a69ffdbfcba8eabf2ca9d384b578e6f28b339c61

Author: Zhenyu Wang <[email protected]>
Date:   Mon Aug 30 16:12:42 2010 +0800

drm/i915: Enable MI_FLUSH on Sandybridge

MI_FLUSH is being deprecated, but still available on Sandybridge. Make sure it's enabled as userspace still uses MI_FLUSH.

It should be removed on Ironlake, and fixed to be 12 on SNB or just removed.

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#7 lore raw
Quoted 15 lines · On Tue, 06 Dec 2011 11:56:07 -0800 Eric Anholt <[email protected]> wrote:
Quoted 3 lines

It's one of those "been there forever" bits. A comment wouldn't hurt, but we should also put in a doc change request (in fact I think I did that but it went into a black hole).

Well, by "been there forever", you mean "introduced in:

commit a69ffdbfcba8eabf2ca9d384b578e6f28b339c61

Author: Zhenyu Wang <[email protected]>
Date:   Mon Aug 30 16:12:42 2010 +0800

drm/i915: Enable MI_FLUSH on Sandybridge

MI_FLUSH is being deprecated, but still available on Sandybridge. Make sure it's enabled as userspace still uses MI_FLUSH.

Quoted 6 lines

It should be removed on Ironlake, and fixed to be 12 on SNB or just removed.

No, "fixed to be 12" would just encourage people to still try it and break their machines. We can safely remove it now that we don't use it anywhere. And we can fix the docs so that if people do want to mess with it they won't try bit 12.

Signature · 1 line
Jesse Barnes, Intel Open Source Technology Center

_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

#8 lore raw
Quoted 21 lines · Jesse Barnes
Quoted 19 lines · On Tue, 06 Dec 2011 11:56:07 -0800 Eric Anholt <[email protected]> wrote:
Quoted 3 lines

It's one of those "been there forever" bits. A comment wouldn't hurt, but we should also put in a doc change request (in fact I think I did that but it went into a black hole).

Well, by "been there forever", you mean "introduced in:

commit a69ffdbfcba8eabf2ca9d384b578e6f28b339c61

Author: Zhenyu Wang <[email protected]>
Date:   Mon Aug 30 16:12:42 2010 +0800

drm/i915: Enable MI_FLUSH on Sandybridge

MI_FLUSH is being deprecated, but still available on Sandybridge. Make sure it's enabled as userspace still uses MI_FLUSH.

Quoted 7 lines
Quoted 2 lines

It should be removed on Ironlake, and fixed to be 12 on SNB or just removed.

No, "fixed to be 12" would just encourage people to still try it and break their machines. We can safely remove it now that we don't use it anywhere. And we can fix the docs so that if people do want to mess with it they won't try bit 12.

If userspace ever relied on this being enabled we cannot just disable it now - it would break the kernel<->userspace abi on snb. You have to wait for 5-10 years, eternity or until the last userspace using this disappeared, whichever comes first ;-)

Cheers, Daniel

Signature · 3 lines
Daniel Vetter
Mail: [email protected]
Mobile: +41 (0)79 365 57 48