[v4] vga_switcheroo: Add helper for deferred probing

Submitted by Lukas Wunner on May 19, 2016, 2:39 p.m.

Details

Message ID 35e4cc1e5c711fa30c3081dc628ce672a0fea276.1463642956.git.lukas@wunner.de
State New
Headers show
Series "vga_switcheroo: Add helper for deferred probing" ( rev: 1 ) in Nouveau

Not browsing as part of any series.

Commit Message

Lukas Wunner May 19, 2016, 2:39 p.m.
So far we've got one condition when DRM drivers need to defer probing
on a dual GPU system and it's coded separately into each of the relevant
drivers. As suggested by Daniel Vetter, deduplicate that code in the
drivers and move it to a new vga_switcheroo helper. This yields better
encapsulation of concepts and lets us add further checks in a central
place. (The existing check pertains to pre-retina MacBook Pros and an
additional check is expected to be needed for retinas.)

v2: This helper could eventually be used by audio clients as well,
    so rephrase kerneldoc to refer to "client" instead of "GPU"
    and move the single existing check in an if block specific
    to PCI_CLASS_DISPLAY_VGA devices. Move documentation on
    that check from kerneldoc to a comment. (Daniel Vetter)

v3: Mandate in kerneldoc that registration of client shall only
    happen after calling this helper. (Daniel Vetter)

v4: Rebase on 412c8f7de011 ("drm/radeon: Return -EPROBE_DEFER when
    amdkfd not loaded")

Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Ben Skeggs <bskeggs@redhat.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Lukas Wunner <lukas@wunner.de>
---
 drivers/gpu/drm/i915/i915_drv.c       | 10 +---------
 drivers/gpu/drm/nouveau/nouveau_drm.c | 10 +---------
 drivers/gpu/drm/radeon/radeon_drv.c   | 10 +---------
 drivers/gpu/vga/vga_switcheroo.c      | 34 ++++++++++++++++++++++++++++++++--
 include/linux/vga_switcheroo.h        |  2 ++
 5 files changed, 37 insertions(+), 29 deletions(-)

Patch hide | download patch | download mbox

diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index dba03c0..61bf5a9 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -35,11 +35,9 @@ 
 #include "i915_trace.h"
 #include "intel_drv.h"
 
-#include <linux/apple-gmux.h>
 #include <linux/console.h>
 #include <linux/module.h>
 #include <linux/pm_runtime.h>
-#include <linux/vgaarb.h>
 #include <linux/vga_switcheroo.h>
 #include <drm/drm_crtc_helper.h>
 
@@ -1030,13 +1028,7 @@  static int i915_pci_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (PCI_FUNC(pdev->devfn))
 		return -ENODEV;
 
-	/*
-	 * apple-gmux is needed on dual GPU MacBook Pro
-	 * to probe the panel if we're the inactive GPU.
-	 */
-	if (IS_ENABLED(CONFIG_VGA_ARB) && IS_ENABLED(CONFIG_VGA_SWITCHEROO) &&
-	    apple_gmux_present() && pdev != vga_default_device() &&
-	    !vga_switcheroo_handler_flags())
+	if (vga_switcheroo_client_probe_defer(pdev))
 		return -EPROBE_DEFER;
 
 	return drm_get_pci_dev(pdev, ent, &driver);
diff --git a/drivers/gpu/drm/nouveau/nouveau_drm.c b/drivers/gpu/drm/nouveau/nouveau_drm.c
index db5c7d0..a0bb1d0 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drm.c
@@ -22,13 +22,11 @@ 
  * Authors: Ben Skeggs
  */
 
-#include <linux/apple-gmux.h>
 #include <linux/console.h>
 #include <linux/delay.h>
 #include <linux/module.h>
 #include <linux/pci.h>
 #include <linux/pm_runtime.h>
-#include <linux/vgaarb.h>
 #include <linux/vga_switcheroo.h>
 
 #include "drmP.h"
@@ -315,13 +313,7 @@  static int nouveau_drm_probe(struct pci_dev *pdev,
 	bool boot = false;
 	int ret;
 
-	/*
-	 * apple-gmux is needed on dual GPU MacBook Pro
-	 * to probe the panel if we're the inactive GPU.
-	 */
-	if (IS_ENABLED(CONFIG_VGA_ARB) && IS_ENABLED(CONFIG_VGA_SWITCHEROO) &&
-	    apple_gmux_present() && pdev != vga_default_device() &&
-	    !vga_switcheroo_handler_flags())
+	if (vga_switcheroo_client_probe_defer(pdev))
 		return -EPROBE_DEFER;
 
 	/* remove conflicting drivers (vesafb, efifb etc) */
diff --git a/drivers/gpu/drm/radeon/radeon_drv.c b/drivers/gpu/drm/radeon/radeon_drv.c
index b55aa74..a455dc7 100644
--- a/drivers/gpu/drm/radeon/radeon_drv.c
+++ b/drivers/gpu/drm/radeon/radeon_drv.c
@@ -34,11 +34,9 @@ 
 #include "radeon_drv.h"
 
 #include <drm/drm_pciids.h>
-#include <linux/apple-gmux.h>
 #include <linux/console.h>
 #include <linux/module.h>
 #include <linux/pm_runtime.h>
-#include <linux/vgaarb.h>
 #include <linux/vga_switcheroo.h>
 #include <drm/drm_gem.h>
 
@@ -340,13 +338,7 @@  static int radeon_pci_probe(struct pci_dev *pdev,
 	if (ret == -EPROBE_DEFER)
 		return ret;
 
-	/*
-	 * apple-gmux is needed on dual GPU MacBook Pro
-	 * to probe the panel if we're the inactive GPU.
-	 */
-	if (IS_ENABLED(CONFIG_VGA_ARB) && IS_ENABLED(CONFIG_VGA_SWITCHEROO) &&
-	    apple_gmux_present() && pdev != vga_default_device() &&
-	    !vga_switcheroo_handler_flags())
+	if (vga_switcheroo_client_probe_defer(pdev))
 		return -EPROBE_DEFER;
 
 	/* Get rid of things like offb */
diff --git a/drivers/gpu/vga/vga_switcheroo.c b/drivers/gpu/vga/vga_switcheroo.c
index cbd7c98..101e14c 100644
--- a/drivers/gpu/vga/vga_switcheroo.c
+++ b/drivers/gpu/vga/vga_switcheroo.c
@@ -30,6 +30,7 @@ 
 
 #define pr_fmt(fmt) "vga_switcheroo: " fmt
 
+#include <linux/apple-gmux.h>
 #include <linux/console.h>
 #include <linux/debugfs.h>
 #include <linux/fb.h>
@@ -308,7 +309,8 @@  static int register_client(struct pci_dev *pdev,
  *
  * Register vga client (GPU). Enable vga_switcheroo if another GPU and a
  * handler have already registered. The power state of the client is assumed
- * to be ON.
+ * to be ON. Beforehand, vga_switcheroo_client_probe_defer() shall be called
+ * to ensure that all prerequisites are met.
  *
  * Return: 0 on success, -ENOMEM on memory allocation error.
  */
@@ -329,7 +331,8 @@  EXPORT_SYMBOL(vga_switcheroo_register_client);
  * @id: client identifier
  *
  * Register audio client (audio device on a GPU). The power state of the
- * client is assumed to be ON.
+ * client is assumed to be ON. Beforehand, vga_switcheroo_client_probe_defer()
+ * shall be called to ensure that all prerequisites are met.
  *
  * Return: 0 on success, -ENOMEM on memory allocation error.
  */
@@ -376,6 +379,33 @@  find_active_client(struct list_head *head)
 }
 
 /**
+ * vga_switcheroo_client_probe_defer() - whether to defer probing a given client
+ * @pdev: client pci device
+ *
+ * Determine whether any prerequisites are not fulfilled to probe a given
+ * client. Drivers shall invoke this early on in their ->probe callback
+ * and return %-EPROBE_DEFER if it evaluates to %true. Thou shalt not
+ * register the client ere thou hast called this.
+ *
+ * Return: %true if probing should be deferred, otherwise %false.
+ */
+bool vga_switcheroo_client_probe_defer(struct pci_dev *pdev)
+{
+	if ((pdev->class >> 8) == PCI_CLASS_DISPLAY_VGA) {
+		/*
+		 * apple-gmux is needed on pre-retina MacBook Pro
+		 * to probe the panel if pdev is the inactive GPU.
+		 */
+		if (apple_gmux_present() && pdev != vga_default_device() &&
+		    !vgasr_priv.handler_flags)
+			return true;
+	}
+
+	return false;
+}
+EXPORT_SYMBOL(vga_switcheroo_client_probe_defer);
+
+/**
  * vga_switcheroo_get_client_state() - obtain power state of a given client
  * @pdev: client pci device
  *
diff --git a/include/linux/vga_switcheroo.h b/include/linux/vga_switcheroo.h
index b39a5f3..960bedb 100644
--- a/include/linux/vga_switcheroo.h
+++ b/include/linux/vga_switcheroo.h
@@ -165,6 +165,7 @@  int vga_switcheroo_unlock_ddc(struct pci_dev *pdev);
 
 int vga_switcheroo_process_delayed_switch(void);
 
+bool vga_switcheroo_client_probe_defer(struct pci_dev *pdev);
 enum vga_switcheroo_state vga_switcheroo_get_client_state(struct pci_dev *dev);
 
 void vga_switcheroo_set_dynamic_switch(struct pci_dev *pdev, enum vga_switcheroo_state dynamic);
@@ -188,6 +189,7 @@  static inline enum vga_switcheroo_handler_flags_t vga_switcheroo_handler_flags(v
 static inline int vga_switcheroo_lock_ddc(struct pci_dev *pdev) { return -ENODEV; }
 static inline int vga_switcheroo_unlock_ddc(struct pci_dev *pdev) { return -ENODEV; }
 static inline int vga_switcheroo_process_delayed_switch(void) { return 0; }
+static inline bool vga_switcheroo_client_probe_defer(struct pci_dev *pdev) { return false; }
 static inline enum vga_switcheroo_state vga_switcheroo_get_client_state(struct pci_dev *dev) { return VGA_SWITCHEROO_ON; }
 
 static inline void vga_switcheroo_set_dynamic_switch(struct pci_dev *pdev, enum vga_switcheroo_state dynamic) {}

Comments

Hi Lukas,

On 19 May 2016 at 15:39, Lukas Wunner <lukas@wunner.de> wrote:

> +bool vga_switcheroo_client_probe_defer(struct pci_dev *pdev)
> +{
> +       if ((pdev->class >> 8) == PCI_CLASS_DISPLAY_VGA) {
Not sure if we want/need this, yet at least. This changes behaviour
which is not what refactoring patches should be doing, right ? if
needed it ought to be a separate patch, imho.

Furthermore on nouveau and AMD (iirc) front, some dual-gpu/optimus
systems use PCI_CLASS_DISPLAY_3D. So if we add DISPLAY_VGA perhaps we
should also check for DISPLAY_3D ?

Regards,
Emil
Hi Emil,

On Fri, May 20, 2016 at 12:41:04AM +0100, Emil Velikov wrote:
> On 19 May 2016 at 15:39, Lukas Wunner <lukas@wunner.de> wrote:
> > +bool vga_switcheroo_client_probe_defer(struct pci_dev *pdev)
> > +{
> > +       if ((pdev->class >> 8) == PCI_CLASS_DISPLAY_VGA) {
> Not sure if we want/need this, yet at least. This changes behaviour
> which is not what refactoring patches should be doing, right ? if
> needed it ought to be a separate patch, imho.

Well, the commit message doesn't claim "no functional change", does it?

Daniel Vetter explicitly wanted the ability to use the helper in
vga_switcheroo audio clients, and those shouldn't run the apple-gmux
test. I think it makes sense to enclose it in the above-quoted if-block
*now* even though it's not needed. Once someone adds a check for an
audio client, chances are they'll forget to add this if-block.

> Furthermore on nouveau and AMD (iirc) front, some dual-gpu/optimus
> systems use PCI_CLASS_DISPLAY_3D. So if we add DISPLAY_VGA perhaps we
> should also check for DISPLAY_3D ?

Fair enough, I've changed it to match PCI_BASE_CLASS_DISPLAY
and sent it out as v5 a few minutes ago.

Thanks,

Lukas
On Sat, 21 May 2016, Lukas Wunner <lukas@wunner.de> wrote:
> Hi Emil,
>
> On Fri, May 20, 2016 at 12:41:04AM +0100, Emil Velikov wrote:
>> On 19 May 2016 at 15:39, Lukas Wunner <lukas@wunner.de> wrote:
>> > +bool vga_switcheroo_client_probe_defer(struct pci_dev *pdev)
>> > +{
>> > +       if ((pdev->class >> 8) == PCI_CLASS_DISPLAY_VGA) {
>> Not sure if we want/need this, yet at least. This changes behaviour
>> which is not what refactoring patches should be doing, right ? if
>> needed it ought to be a separate patch, imho.
>
> Well, the commit message doesn't claim "no functional change", does it?
>
> Daniel Vetter explicitly wanted the ability to use the helper in
> vga_switcheroo audio clients, and those shouldn't run the apple-gmux
> test. I think it makes sense to enclose it in the above-quoted if-block
> *now* even though it's not needed. Once someone adds a check for an
> audio client, chances are they'll forget to add this if-block.

Please always keep functional changes separate from non-functional
refactoring. Combining them into one just makes everyone's life
harder. This may seem like a simple case here, but you should get into
the habit and stick to it.

If you do a non-functional refactoring change, it's easy for people to
review that change. I could've reviewed your patch in a few minutes;
that's what I was in fact going to do until I stumbled on the functional
change. Maybe I don't have the time to dig all the implications in that
change right now. Bummer.

If you do a non-functional refactoring change, and claim it's
non-functional in the commit message, it'll help anyone landing on that
commit as a result of bisecting a regression. Either the bisect was
bogus, or your patch was unintentionally functional after all. Either
should be easier to verify than a combined change.

If you do the functional change separately, and that regresses, you
don't have to throw out perfectly good refactoring changes when you
revert the functional change. Or maybe the refactoring change doesn't
revert anymore, and you'll have to do a more involved fix on top. It's
easy to do reverts of clearly regressing commits, even by people who
don't fully understand everything about the commit. Anything more is
harder. Not rocket science, but it just adds up to the burden of the
already swamped maintainers.

Usually, like here, you should do the non-functional refactoring first,
and then the functional change on top. The notable exception to this is
when the functional change is a fix, especially of the kind that may
need to get backported. Then you should do the fix first, even if it's
harder and more verbose that way, because you don't want to add a stable
backport dependency on a refactoring commit (which may exceed stable
patch limits) and the stable maintainers frown upon manually backported
patches.

> Fair enough, I've changed it to match PCI_BASE_CLASS_DISPLAY
> and sent it out as v5 a few minutes ago.

Please make that a v6 and split up the patch.


Thanks,
Jani.