[PATCH v5 0/2] media: hantro: fix runtime PM resource handling
Tharit Tangkijwanichakul <[email protected]>
| Newsgroups | org.infradead.lists.linux-rockchip,dev.linux.lists.linux-kernel-mentees,org.infradead.lists.linux-arm-kernel,org.kernel.vger.linux-kernel,org.kernel.vger.linux-media |
|---|---|
| Message-ID | <[email protected]> |
The Hantro device_run() path acquires a runtime PM reference before invoking the codec-specific run callback. Failure paths can leave the runtime PM reference and enabled clocks held. Patch 1 moves clock enable and disable operations into the runtime PM callbacks and releases the runtime PM reference when device_run() fails. It retains the existing CONFIG_PM conditional so that the patch remains independently buildable. Patch 2 removes the explicit CONFIG_PM conditional, defines the PM operations with DEFINE_RUNTIME_DEV_PM_OPS(), and uses pm_ptr() when assigning the PM operations to the platform driver. Changes in v5: - Split the runtime PM changes into two patches. - Move clock management into the runtime PM callbacks. - Release the runtime PM reference on device_run() failure. - Retain the CONFIG_PM conditional in patch 1. - Remove the CONFIG_PM conditional in patch 2 using DEFINE_RUNTIME_DEV_PM_OPS() and pm_ptr(). v4: https://lore.kernel.org/linux-media/[email protected] Tharit Tangkijwanichakul (2): media: hantro: release runtime resources when device_run fails media: hantro: use DEFINE_RUNTIME_DEV_PM_OPS .../media/platform/verisilicon/hantro_drv.c | 75 ++++++++++--------- 1 file changed, 41 insertions(+), 34 deletions(-) --- Tested on a Rockchip RK3588 (Rock 5B) board with Fluster: H.264 (JVT-AVC_V1): 129/135, unchanged MPEG-2 (MPEG2_VIDEO-MAIN): 23/43, unchanged VP8 (VP8-TEST-VECTORS): 61/61, unchanged base-commit: dc59e4fea9d83f03bad6bddf3fa2e52491777482 -- 2.47.3 _______________________________________________ Linux-rockchip mailing list [email protected] http://lists.infradead.org/mailman/listinfo/linux-rockchip