Skip to contentSkip to conversation

drm/i915: Allow the UMD to configure their own power clock state

RFC In review Oscar Mateo 7 participants 10 messages May 2, 2017 – May 11, 2017

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

RFC 1/1 drm/i915: Allow the UMD to configure their own power clock state

This allows userspace to shutdown slices at will for performance/power reasons (because it doesn't have a use for more slices).

Cc Dmitry Rogozhkin <[email protected]>
Cc Chris Wilson <[email protected]>
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/gpu/drm/i915/intel_engine_cs.c +11 -0
index 402769d..17ff88d 100644
@@ -628,6 +628,7 @@ static int wa_ring_whitelist_reg(struct intel_engine_cs *engine,
 static int gen8_init_workarounds(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
+	int ret;
 
 	WA_SET_BIT_MASKED(INSTPM, INSTPM_FORCE_ORDERING);
 
@@ -673,6 +674,11 @@ static int gen8_init_workarounds(struct intel_engine_cs *engine)
 			    GEN6_WIZ_HASHING_MASK,
 			    GEN6_WIZ_HASHING_16x4);
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+
 	return 0;
 }
 
@@ -841,6 +847,11 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+
 	return 0;
 }
 
Signature · 6 lines
1.9.1

_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#1 lore raw

✓ Fi.CI.BAT: success for drm/i915: Allow the UMD to configure their own power clock state

== Series Details ==

== Series Details ==

Series: drm/i915: Allow the UMD to configure their own power clock state
URL : https://patchwork.freedesktop.org/series/23846/
State : success

== Summary ==

Series 23846v1 drm/i915: Allow the UMD to configure their own power clock state
https://patchwork.freedesktop.org/api/1.0/series/23846/revisions/1/mbox/

Test drv_module_reload:

        Subgroup basic-reload-final:
                dmesg-warn -> PASS       (fi-skl-6770hq) fdo#100248

fdo#100248 https://bugs.freedesktop.org/show_bug.cgi?id=100248

fi-bdw-5557u     total:278  pass:267  dwarn:0   dfail:0   fail:0   skip:11  time:437s
fi-bdw-gvtdvm    total:278  pass:256  dwarn:8   dfail:0   fail:0   skip:14  time:427s
fi-bsw-n3050     total:278  pass:242  dwarn:0   dfail:0   fail:0   skip:36  time:581s
fi-bxt-j4205     total:278  pass:259  dwarn:0   dfail:0   fail:0   skip:19  time:508s
fi-bxt-t5700     total:278  pass:258  dwarn:0   dfail:0   fail:0   skip:20  time:562s
fi-byt-j1900     total:278  pass:254  dwarn:0   dfail:0   fail:0   skip:24  time:491s
fi-byt-n2820     total:278  pass:250  dwarn:0   dfail:0   fail:0   skip:28  time:491s
fi-hsw-4770      total:278  pass:262  dwarn:0   dfail:0   fail:0   skip:16  time:409s
fi-hsw-4770r     total:278  pass:262  dwarn:0   dfail:0   fail:0   skip:16  time:405s
fi-ilk-650       total:278  pass:228  dwarn:0   dfail:0   fail:0   skip:50  time:421s
fi-ivb-3520m     total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:493s
fi-ivb-3770      total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:464s
fi-kbl-7500u     total:278  pass:260  dwarn:0   dfail:0   fail:0   skip:18  time:458s
fi-kbl-7560u     total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10  time:568s
fi-skl-6260u     total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10  time:450s
fi-skl-6700hq    total:278  pass:261  dwarn:0   dfail:0   fail:0   skip:17  time:579s
fi-skl-6700k     total:278  pass:256  dwarn:4   dfail:0   fail:0   skip:18  time:461s
fi-skl-6770hq    total:278  pass:268  dwarn:0   dfail:0   fail:0   skip:10  time:490s
fi-skl-gvtdvm    total:278  pass:265  dwarn:0   dfail:0   fail:0   skip:13  time:431s
fi-snb-2520m     total:278  pass:250  dwarn:0   dfail:0   fail:0   skip:28  time:531s
fi-snb-2600      total:278  pass:249  dwarn:0   dfail:0   fail:0   skip:29  time:400s

7f027554339633cb4e6a2d3974510aaef1a9e24c drm-tip: 2017y-05m-02d-18h-48m-28s UTC integration manifest 7a90a9b drm/i915: Allow the UMD to configure their own power clock state

== Logs ==

For more details see: https://intel-gfx-ci.01.org/CI/Patchwork_4602/
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx

maintainer
#2 lore raw
In reply to Oscar Mateo
Quoted 6 lines · Oscar Mateo

This allows userspace to shutdown slices at will for performance/power reasons (because it doesn't have a use for more slices).

Cc Dmitry Rogozhkin <[email protected]>
Cc Chris Wilson <[email protected]>

<SNIP>

Quoted 16 lines
drivers/gpu/drm/i915/intel_engine_cs.c +0 -0
@@ -673,6 +674,11 @@ static int gen8_init_workarounds(struct intel_engine_cs *engine)
  			    GEN6_WIZ_HASHING_MASK,
  			    GEN6_WIZ_HASHING_16x4);

drivers/gpu/drm/i915/intel_engine_cs.c +5 -0
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+

  return 0;
 }

drivers/gpu/drm/i915/intel_engine_cs.c +0 -0
@@ -841,6 +847,11 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine)
  	if (ret)
  		return ret;

drivers/gpu/drm/i915/intel_engine_cs.c static int gen8_init_workarounds(struct intel_engine_cs *engine) · line 680

Quoted 4 lines
drivers/gpu/drm/i915/intel_engine_cs.c +4 -0
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;

This is not a workaround, so it should be part of the cmd parser and have an userspace.

Regards, Joonas

Signature · 7 lines
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#3 lore raw
Quoted 11 lines · Joonas Lahtinen

drivers/gpu/drm/i915/intel_engine_cs.c static int gen9_init_workarounds(struct intel_engine_cs *engine) · line 853

Quoted 8 lines
drivers/gpu/drm/i915/intel_engine_cs.c +4 -0
@@ -841,6 +847,11 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine)
      if (ret)
              return ret;
 
+     /* Allow the UMD to configure their own power clock state */
+     ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+     if (ret)
+             return ret;

This is not a workaround, so it should be part of the cmd parser and have an userspace.

Yeah, the "This allows userspace ..." start of the commit message should have been a dead giveaway that we do indeed need the userspace patch for this too.
-Daniel

Signature · 7 lines
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#4 lore raw
Quoted 6 lines · Daniel Vetter
Quoted 4 lines · Joonas Lahtinen
Quoted 4 lines
drivers/gpu/drm/i915/intel_engine_cs.c +0 -0
@@ -841,6 +847,11 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine)
       if (ret)
               return ret;
 
Quoted 6 lines
Quoted 6 lines

drivers/gpu/drm/i915/intel_engine_cs.c static int gen8_init_workarounds(struct intel_engine_cs *engine) · line 680

Quoted 4 lines
drivers/gpu/drm/i915/intel_engine_cs.c +4 -0
+     /* Allow the UMD to configure their own power clock state */
+     ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+     if (ret)
+             return ret;

This is not a workaround, so it should be part of the cmd parser and have an userspace.

cmd parser? This is Gen8+, so the cmd parser shouldn't even be active. Or am I missing something?

Quoted 4 lines

Yeah, the "This allows userspace ..." start of the commit message should have been a dead giveaway that we do indeed need the userspace patch for this too.
-Daniel

Yes, I got the memo :)
Dmitry has opened a ticket against intel-vaapi-driver (https://github.com/01org/intel-vaapi-driver/issues/152) so hopefully we will have a userspace soon.

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

maintainer
#5 lore raw
Quoted 23 lines · Oscar Mateo
Quoted 10 lines · Daniel Vetter
Quoted 7 lines · Joonas Lahtinen
Quoted 7 lines
drivers/gpu/drm/i915/intel_engine_cs.c +0 -0
@@ -841,6 +847,11 @@ static int gen9_init_workarounds(struct
 intel_engine_cs *engine)
        if (ret)
                return ret;
 
Quoted 9 lines
Quoted 9 lines

Review static int gen9_init_workarounds(struct · line 856

Quoted 6 lines
(quoted diff) +4 -0
+     /* Allow the UMD to configure their own power clock state
 */
+     ret = wa_ring_whitelist_reg(engine,
 GEN8_R_PWR_CLK_STATE);
+     if (ret)
+             return ret;

This is not a workaround, so it should be part of the cmd parser and
have an userspace.

Quoted 3 lines

cmd parser? This is Gen8+, so the cmd parser shouldn't even be active.
Or am I missing something?

That's correct, my bad.

Quoted 8 lines
Quoted 7 lines

Yeah, the "This allows userspace ..." start of the commit message should have been a dead giveaway that we do indeed need the userspace
patch for this too.
-Daniel

Quoted 5 lines

Yes, I got the memo :)
Dmitry has opened a ticket against intel-vaapi-driver
(https://github.com/01org/intel-vaapi-driver/issues/152) so hopefully we
will have a userspace soon.

So with the userspace, we can merge.

Regards, Joonas

Signature · 7 lines
Joonas Lahtinen
Open Source Technology Center
Intel Corporation
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#6 lore raw
In reply to Oscar Mateo
Quoted 23 lines · Oscar Mateo

This allows userspace to shutdown slices at will for performance/power reasons (because it doesn't have a use for more slices).

Cc Dmitry Rogozhkin <[email protected]>
Cc Chris Wilson <[email protected]>
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/gpu/drm/i915/intel_engine_cs.c +1 -0
index 402769d..17ff88d 100644
@@ -628,6 +628,7 @@ static int wa_ring_whitelist_reg(struct intel_engine_cs *engine,
 static int gen8_init_workarounds(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
+	int ret;
 
 	WA_SET_BIT_MASKED(INSTPM, INSTPM_FORCE_ORDERING);
 

drivers/gpu/drm/i915/intel_engine_cs.c static int gen8_init_workarounds(struct intel_engine_cs *engine) · line 681

Quoted 8 lines
drivers/gpu/drm/i915/intel_engine_cs.c +5 -0
@@ -673,6 +674,11 @@ static int gen8_init_workarounds(struct intel_engine_cs *engine)
 			    GEN6_WIZ_HASHING_MASK,
 			    GEN6_WIZ_HASHING_16x4);
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+

How will the different userspace clients coordinate with the slice state? I guess in another words, is this part of the context?

-Mika

Quoted 22 lines
drivers/gpu/drm/i915/intel_engine_cs.c +5 -0
 	return 0;
 }
 
@@ -841,6 +847,11 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+
 	return 0;
 }
 
Signature · 6 lines
1.9.1

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

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

#7 lore raw
Quoted 37 lines · Mika Kuoppala
Quoted 23 lines · Oscar Mateo

This allows userspace to shutdown slices at will for performance/power reasons (because it doesn't have a use for more slices).

Cc Dmitry Rogozhkin <[email protected]>
Cc Chris Wilson <[email protected]>
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/gpu/drm/i915/intel_engine_cs.c +1 -0
index 402769d..17ff88d 100644
@@ -628,6 +628,7 @@ static int wa_ring_whitelist_reg(struct intel_engine_cs *engine,
 static int gen8_init_workarounds(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
+	int ret;
 
 	WA_SET_BIT_MASKED(INSTPM, INSTPM_FORCE_ORDERING);
 

drivers/gpu/drm/i915/intel_engine_cs.c static int gen8_init_workarounds(struct intel_engine_cs *engine) · line 681

Quoted 8 lines
drivers/gpu/drm/i915/intel_engine_cs.c +5 -0
@@ -673,6 +674,11 @@ static int gen8_init_workarounds(struct intel_engine_cs *engine)
 			    GEN6_WIZ_HASHING_MASK,
 			    GEN6_WIZ_HASHING_16x4);
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+

How will the different userspace clients coordinate with the slice state? I guess in another words, is this part of the context?

Quoted 2 lines

-Mika

Spec says (see IHD-OS-SKL-Vol2c-05.16 page 614):
"This register must be initialized correctly when the context is submitted for the first time. This register is context save/restored as part of Exec-List context image in both Exec-List and Ring-Buffer mode of scheduling."

So we're good, right?

Except there's also:
"This register must not be programmed using MI_LOAD_REGISTER_IMM command in ring buffer or in batch buffer, however programming "NON - SLM Indication" field through MI_LOAD_REGISTER_IMM is an exception defined below. If a need arises to change the render configuration for a context being executed in HW, Scheduler must preempt the context and update the desired render configuration in the logical render context image in memory and resubmit the context"

So... are we sure that we're fine with giving userspace control over this register? If so - it would still be a good idea to mention that the info in the spec is bogus (commit message or just a comment).

-Michał

Quoted 19 lines
Quoted 19 lines
drivers/gpu/drm/i915/intel_engine_cs.c +5 -0
 	return 0;
 }
 
@@ -841,6 +847,11 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine)
 	if (ret)
 		return ret;
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+
 	return 0;
 }
 
Signature · 1 line
1.9.1
Quoted 8 lines
Quoted 4 lines

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

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

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

#8 lore raw
Quoted 56 lines · Michał Winiarski
Quoted 37 lines · Mika Kuoppala
Quoted 23 lines · Oscar Mateo

This allows userspace to shutdown slices at will for performance/power reasons (because it doesn't have a use for more slices).

Cc Dmitry Rogozhkin <[email protected]>
Cc Chris Wilson <[email protected]>
---
 drivers/gpu/drm/i915/intel_engine_cs.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/gpu/drm/i915/intel_engine_cs.c +1 -0
index 402769d..17ff88d 100644
@@ -628,6 +628,7 @@ static int wa_ring_whitelist_reg(struct intel_engine_cs *engine,
 static int gen8_init_workarounds(struct intel_engine_cs *engine)
 {
 	struct drm_i915_private *dev_priv = engine->i915;
+	int ret;
 
 	WA_SET_BIT_MASKED(INSTPM, INSTPM_FORCE_ORDERING);
 

drivers/gpu/drm/i915/intel_engine_cs.c static int gen8_init_workarounds(struct intel_engine_cs *engine) · line 681

Quoted 8 lines
drivers/gpu/drm/i915/intel_engine_cs.c +5 -0
@@ -673,6 +674,11 @@ static int gen8_init_workarounds(struct intel_engine_cs *engine)
 			    GEN6_WIZ_HASHING_MASK,
 			    GEN6_WIZ_HASHING_16x4);
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+

How will the different userspace clients coordinate with the slice state? I guess in another words, is this part of the context?

Quoted 2 lines

-Mika

Spec says (see IHD-OS-SKL-Vol2c-05.16 page 614):
"This register must be initialized correctly when the context is submitted for the first time. This register is context save/restored as part of Exec-List context image in both Exec-List and Ring-Buffer mode of scheduling."

So we're good, right?

Except there's also:
"This register must not be programmed using MI_LOAD_REGISTER_IMM command in ring buffer or in batch buffer, however programming "NON - SLM Indication" field through MI_LOAD_REGISTER_IMM is an exception defined below. If a need arises to change the render configuration for a context being executed in HW, Scheduler must preempt the context and update the desired render configuration in the logical render context image in memory and resubmit the context"

Quoted 3 lines

So... are we sure that we're fine with giving userspace control over this register? If so - it would still be a good idea to mention that the info in the spec is bogus (commit message or just a comment).

Doesn't sound like it. So we are back to doing the preempt and changing the context image as suggested in the other patch. Still we have the issue that OA config conflicts with changing the slice eu config. -Chris

Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
#9 lore raw
In reply to Michał Winiarski
Quoted 37 lines · Michał Winiarski
Quoted 36 lines · Mika Kuoppala
Quoted 23 lines · Oscar Mateo

This allows userspace to shutdown slices at will for performance/power reasons (because it doesn't have a use for more slices).

Cc Dmitry Rogozhkin <[email protected]>
Cc Chris Wilson <[email protected]>
---
  drivers/gpu/drm/i915/intel_engine_cs.c | 11 +++++++++++
  1 file changed, 11 insertions(+)
drivers/gpu/drm/i915/intel_engine_cs.c +1 -0
index 402769d..17ff88d 100644
@@ -628,6 +628,7 @@ static int wa_ring_whitelist_reg(struct intel_engine_cs *engine,
  static int gen8_init_workarounds(struct intel_engine_cs *engine)
  {
  	struct drm_i915_private *dev_priv = engine->i915;
+	int ret;
 
  	WA_SET_BIT_MASKED(INSTPM, INSTPM_FORCE_ORDERING);
 

drivers/gpu/drm/i915/intel_engine_cs.c static int gen8_init_workarounds(struct intel_engine_cs *engine) · line 681

Quoted 8 lines
drivers/gpu/drm/i915/intel_engine_cs.c +5 -0
@@ -673,6 +674,11 @@ static int gen8_init_workarounds(struct intel_engine_cs *engine)
  			    GEN6_WIZ_HASHING_MASK,
  			    GEN6_WIZ_HASHING_16x4);
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+

How will the different userspace clients coordinate with the slice state? I guess in another words, is this part of the context?

Quoted 1 line
Quoted 1 line

-Mika

As Michal said, this is context-saved/restored, so no need to coordinate between clients.

Quoted 6 lines

Spec says (see IHD-OS-SKL-Vol2c-05.16 page 614):
"This register must be initialized correctly when the context is submitted for the first time. This register is context save/restored as part of Exec-List context image in both Exec-List and Ring-Buffer mode of scheduling."

So we're good, right?

Ditto :)

Quoted 9 lines

Except there's also:
"This register must not be programmed using MI_LOAD_REGISTER_IMM command in ring buffer or in batch buffer, however programming "NON - SLM Indication" field through MI_LOAD_REGISTER_IMM is an exception defined below. If a need arises to change the render configuration for a context being executed in HW, Scheduler must preempt the context and update the desired render configuration in the logical render context image in memory and resubmit the context"

Quoted 3 lines

So... are we sure that we're fine with giving userspace control over this register? If so - it would still be a good idea to mention that the info in the spec is bogus (commit message or just a comment).

This approach has been double-checked with HW and they believe there should be no problem, as long as the update is done as:

- PIPECONTROL - Stalling flish, flush all caches (color, depth, DC$)
- LOAD_REGISTER_IMMEDIATE - R_PWR_CLK_STATE
- Reprogram complete state

I'm trying to get them to change the BSpec accordingly.

Quoted 18 lines
Quoted 18 lines
Quoted 18 lines
drivers/gpu/drm/i915/intel_engine_cs.c +5 -0
  	return 0;
  }
 
@@ -841,6 +847,11 @@ static int gen9_init_workarounds(struct intel_engine_cs *engine)
  	if (ret)
  		return ret;
 
+	/* Allow the UMD to configure their own power clock state */
+	ret = wa_ring_whitelist_reg(engine, GEN8_R_PWR_CLK_STATE);
+	if (ret)
+		return ret;
+
  	return 0;
  }
 
Signature · 1 line
1.9.1
Quoted 10 lines
Quoted 10 lines
Quoted 4 lines

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

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

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