* [PATCH v2 1/2] ALSA: hda: intel: Release i915 power also when delayed probe fails
@ 2026-11-20 15:09 Kai Vehmanen
2026-11-20 15:09 ` [PATCH v2 2/2] ALSA: hda: intel: Remove unused free_on_error from azx_probe_continue() Kai Vehmanen
2026-11-22 8:24 ` [PATCH v2 1/2] ALSA: hda: intel: Release i915 power also when delayed probe fails Takashi Iwai
0 siblings, 2 replies; 3+ messages in thread
From: Kai Vehmanen @ 2026-11-20 15:09 UTC (permalink / raw)
To: linux-sound, tiwai, Peter Ujfalusi
Cc: intel-gfx, kai.vehmanen, Michal Grzelak
Commit fbaccb1b19c7 ("ALSA: hda: intel: Don't free card in the probe
work") stopped freeing the card from azx_probe_work() when the delayed
probe fails, and only disconnects it, leaving the free to azx_remove().
This avoids the double free, but the resources that were released by
the card free via azx_free() are now kept until the PCI device is
unbound. In particular the display power reference stay held, blocking
runtime suspend for i915/xe.
Release the display power in the azx_probe_continue() work in all cases.
This power reference is only taken for the purpose of the display codec
probe. If the display codec probe was successful, it will have taken its
own reference. In case of systems with the HDA controller on GPU side
(hda->need_i915_power), functionality is same as before - we have to
keep power.
Fixes: fbaccb1b19c7 ("ALSA: hda: intel: Don't free card in the probe work")
Reported-by: Michal Grzelak <michal.grzelak@intel.com>
Link: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/17751
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
---
Notes for V2:
- Reduced the scope of change based sashikobot comments. In this version,
only drop the display power reference taken for probing and do the change
in in probe_continue(). Leave all the rest to azx_remove().
- Add a second patch to remove unused code left after commit fbaccb1b19c7
sound/hda/controllers/intel.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/sound/hda/controllers/intel.c b/sound/hda/controllers/intel.c
index 5a7dd69955d4..1f754f04eea8 100644
--- a/sound/hda/controllers/intel.c
+++ b/sound/hda/controllers/intel.c
@@ -2452,6 +2452,9 @@ static int azx_probe_continue(struct azx *chip, bool free_on_error)
}
out_free:
+ if (!hda->need_i915_power)
+ display_power(chip, false);
+
if (err < 0) {
if (hda->vga_switcheroo_registered)
dev_warn(&pci->dev, "GPU sound probed, but not operational: please add a quirk to driver_denylist\n");
@@ -2462,8 +2465,6 @@ static int azx_probe_continue(struct azx *chip, bool free_on_error)
return err;
}
- if (!hda->need_i915_power)
- display_power(chip, false);
complete_all(&hda->probe_wait);
to_hda_bus(bus)->bus_probing = 0;
hda->probe_retry = 0;
base-commit: 1afd6b7547b2ff0824e91fdf269bbd089f767a5d
--
2.43.0
^ permalink raw reply related [flat|nested] 3+ messages in thread
* [PATCH v2 2/2] ALSA: hda: intel: Remove unused free_on_error from azx_probe_continue()
2026-11-20 15:09 [PATCH v2 1/2] ALSA: hda: intel: Release i915 power also when delayed probe fails Kai Vehmanen
@ 2026-11-20 15:09 ` Kai Vehmanen
2026-11-22 8:24 ` [PATCH v2 1/2] ALSA: hda: intel: Release i915 power also when delayed probe fails Takashi Iwai
1 sibling, 0 replies; 3+ messages in thread
From: Kai Vehmanen @ 2026-11-20 15:09 UTC (permalink / raw)
To: linux-sound, tiwai, Peter Ujfalusi; +Cc: intel-gfx, kai.vehmanen
Commit fbaccb1b19c7 ("ALSA: hda: intel: Don't free card in the probe
work") changed azx_probe_work() to pass false for the free_on_error
argument of azx_probe_continue(), and left freeing the card to
azx_remove(). That was the only caller that passed true, so the
argument is now always false and the code freeing the card on error is
dead.
Remove the argument and the dead code.
No functional change.
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
---
sound/hda/controllers/intel.c | 12 ++++--------
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/sound/hda/controllers/intel.c b/sound/hda/controllers/intel.c
index 1f754f04eea8..acca8bb9ef0b 100644
--- a/sound/hda/controllers/intel.c
+++ b/sound/hda/controllers/intel.c
@@ -1184,7 +1184,7 @@ static const struct dev_pm_ops azx_pm = {
};
-static int azx_probe_continue(struct azx *chip, bool free_on_error);
+static int azx_probe_continue(struct azx *chip);
#ifdef SUPPORT_VGA_SWITCHEROO
static struct pci_dev *get_bound_vga(struct pci_dev *pci);
@@ -1211,7 +1211,7 @@ static void azx_vs_set_state(struct pci_dev *pci,
if (!disabled) {
dev_info(chip->card->dev,
"Start delayed initialization\n");
- if (azx_probe_continue(chip, false) < 0) {
+ if (azx_probe_continue(chip) < 0) {
hda->init_failed = 1;
dev_err(chip->card->dev, "initialization error\n");
}
@@ -1739,7 +1739,7 @@ static void azx_probe_work(struct work_struct *work)
{
struct hda_intel *hda = container_of(work, struct hda_intel, probe_work.work);
- if (azx_probe_continue(&hda->chip, false) < 0) {
+ if (azx_probe_continue(&hda->chip) < 0) {
hda->init_failed = 1;
complete_all(&hda->probe_wait);
snd_card_disconnect(hda->chip.card);
@@ -2362,7 +2362,7 @@ static const unsigned int azx_max_codecs[AZX_NUM_DRIVERS] = {
[AZX_DRIVER_TERA] = 1,
};
-static int azx_probe_continue(struct azx *chip, bool free_on_error)
+static int azx_probe_continue(struct azx *chip)
{
struct hda_intel *hda = container_of(chip, struct hda_intel, chip);
struct hdac_bus *bus = azx_bus(chip);
@@ -2458,10 +2458,6 @@ static int azx_probe_continue(struct azx *chip, bool free_on_error)
if (err < 0) {
if (hda->vga_switcheroo_registered)
dev_warn(&pci->dev, "GPU sound probed, but not operational: please add a quirk to driver_denylist\n");
- if (free_on_error) {
- pci_set_drvdata(pci, NULL);
- snd_card_free(chip->card);
- }
return err;
}
--
2.43.0
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH v2 1/2] ALSA: hda: intel: Release i915 power also when delayed probe fails
2026-11-20 15:09 [PATCH v2 1/2] ALSA: hda: intel: Release i915 power also when delayed probe fails Kai Vehmanen
2026-11-20 15:09 ` [PATCH v2 2/2] ALSA: hda: intel: Remove unused free_on_error from azx_probe_continue() Kai Vehmanen
@ 2026-11-22 8:24 ` Takashi Iwai
1 sibling, 0 replies; 3+ messages in thread
From: Takashi Iwai @ 2026-11-22 8:24 UTC (permalink / raw)
To: Kai Vehmanen
Cc: linux-sound, tiwai, Peter Ujfalusi, intel-gfx, Michal Grzelak
On Fri, 20 Nov 2026 16:09:36 +0100,
Kai Vehmanen wrote:
>
> Commit fbaccb1b19c7 ("ALSA: hda: intel: Don't free card in the probe
> work") stopped freeing the card from azx_probe_work() when the delayed
> probe fails, and only disconnects it, leaving the free to azx_remove().
> This avoids the double free, but the resources that were released by
> the card free via azx_free() are now kept until the PCI device is
> unbound. In particular the display power reference stay held, blocking
> runtime suspend for i915/xe.
>
> Release the display power in the azx_probe_continue() work in all cases.
> This power reference is only taken for the purpose of the display codec
> probe. If the display codec probe was successful, it will have taken its
> own reference. In case of systems with the HDA controller on GPU side
> (hda->need_i915_power), functionality is same as before - we have to
> keep power.
>
> Fixes: fbaccb1b19c7 ("ALSA: hda: intel: Don't free card in the probe work")
> Reported-by: Michal Grzelak <michal.grzelak@intel.com>
> Link: https://gitlab.freedesktop.org/drm/i915/kernel/-/work_items/17751
> Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Applied now. Thanks.
Takashi
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-11-22 8:24 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-20 15:09 [PATCH v2 1/2] ALSA: hda: intel: Release i915 power also when delayed probe fails Kai Vehmanen
2026-11-20 15:09 ` [PATCH v2 2/2] ALSA: hda: intel: Remove unused free_on_error from azx_probe_continue() Kai Vehmanen
2026-11-22 8:24 ` [PATCH v2 1/2] ALSA: hda: intel: Release i915 power also when delayed probe fails Takashi Iwai
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox