linux-media.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Sergey Senozhatsky <senozhatsky@chromium.org>
To: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: Bryan O'Donoghue <bryan.odonoghue@linaro.org>,
	Stanimir Varbanov <stanimir.k.varbanov@gmail.com>,
	Vikash Garodia <quic_vgarodia@quicinc.com>,
	linux-media@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCHv2 2/2] media: venus: sync with threaded IRQ during inst destruction
Date: Thu, 24 Oct 2024 20:36:49 +0900	[thread overview]
Message-ID: <20241024113649.GO1279924@google.com> (raw)
In-Reply-To: <20241024100857.GN1279924@google.com>

On (24/10/24 19:08), Sergey Senozhatsky wrote:
> So, we need to
> - export a couple of symbols
> - include vdec header in core
[..]
> +++ b/drivers/media/platform/qcom/venus/core.c
> @@ -500,6 +500,31 @@ static __maybe_unused int venus_runtime_suspend(struct device *dev)
>  	return ret;
>  }
>  
> +void venus_close_common(struct venus_inst *inst)
> +{
> +	/*
> +	 * First, remove the inst from the ->instances list, so that
> +	 * to_instance() will return NULL.
> +	 */
> +	hfi_session_destroy(inst);
> +	/*
> +	 * Second, make sure we don't have IRQ/IRQ-thread currently running
> +	 * or pending execution, which would race with the inst destruction.
> +	 */
> +	synchronize_irq(inst->core->irq);
> +
> +	v4l2_m2m_ctx_release(inst->m2m_ctx);
> +	v4l2_m2m_release(inst->m2m_dev);
> +	ida_destroy(&inst->dpb_ids);
> +	v4l2_fh_del(&inst->fh);
> +	v4l2_fh_exit(&inst->fh);
> +	vdec_ctrl_deinit(inst);
> +
> +	mutex_destroy(&inst->lock);
> +	mutex_destroy(&inst->ctx_q_lock);
> +}
> +EXPORT_SYMBOL_GPL(venus_close_common);
[..]
> +++ b/drivers/media/platform/qcom/venus/core.h
> @@ -17,6 +17,7 @@
>  #include "hfi.h"
>  #include "hfi_platform.h"
>  #include "hfi_helper.h"
> +#include "vdec.h"
>  
>  #define VDBGL	"VenusLow : "
>  #define VDBGM	"VenusMed : "
> @@ -569,4 +570,6 @@ is_fw_rev_or_older(struct venus_core *core, u32 vmajor, u32 vminor, u32 vrev)
>  		(core)->venus_ver.minor == vminor &&
>  		(core)->venus_ver.rev <= vrev);
>  }
> +
> +void venus_close_common(struct venus_inst *inst);
>  #endif
[..]
> +++ b/drivers/media/platform/qcom/venus/vdec_ctrls.c
> @@ -192,3 +192,4 @@ void vdec_ctrl_deinit(struct venus_inst *inst)
>  {
>  	v4l2_ctrl_handler_free(&inst->ctrl_handler);
>  }
> +EXPORT_SYMBOL_GPL(vdec_ctrl_deinit);

Would have been entirely too simple had I compile-tested
my patches, wouldn't it?

	depmod: ERROR: Cycle detected: venus_core -> venus_dec -> venus_core


vdec_ctrl_deinit() is a v4l2_ctrl_handler_free() wrapper, a one-liner,
so I turned into a static inline.

Compiles with the addition of:

---

diff --git a/drivers/media/platform/qcom/venus/vdec.h b/drivers/media/platform/qcom/venus/vdec.h
index 6b262d0bf561..2687255b1616 100644
--- a/drivers/media/platform/qcom/venus/vdec.h
+++ b/drivers/media/platform/qcom/venus/vdec.h
@@ -6,9 +6,14 @@
 #ifndef __VENUS_VDEC_H__
 #define __VENUS_VDEC_H__
 
+#include <media/v4l2-ctrls.h>
+
 struct venus_inst;
 
 int vdec_ctrl_init(struct venus_inst *inst);
-void vdec_ctrl_deinit(struct venus_inst *inst);
+static inline void vdec_ctrl_deinit(struct venus_inst *inst)
+{
+       v4l2_ctrl_handler_free(&inst->ctrl_handler);
+}
 
 #endif
diff --git a/drivers/media/platform/qcom/venus/vdec_ctrls.c b/drivers/media/platform/qcom/venus/vdec_ctrls.c
index 2b6b2eee619c..fa034a7fdbed 100644
--- a/drivers/media/platform/qcom/venus/vdec_ctrls.c
+++ b/drivers/media/platform/qcom/venus/vdec_ctrls.c
@@ -4,7 +4,6 @@
  * Copyright (C) 2017 Linaro Ltd.
  */
 #include <linux/types.h>
-#include <media/v4l2-ctrls.h>
 
 #include "core.h"
 #include "helpers.h"
@@ -187,9 +186,3 @@ int vdec_ctrl_init(struct venus_inst *inst)
 
        return 0;
 }
-
-void vdec_ctrl_deinit(struct venus_inst *inst)
-{
-       v4l2_ctrl_handler_free(&inst->ctrl_handler);
-}

      parent reply	other threads:[~2024-10-24 11:36 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-10-24  6:16 [PATCHv2 0/2] media: venus: close() fixes Sergey Senozhatsky
2024-10-24  6:16 ` [PATCHv2 1/2] media: venus: fix enc/dec destruction order Sergey Senozhatsky
2024-10-24  8:54   ` Bryan O'Donoghue
2024-10-24  6:16 ` [PATCHv2 2/2] media: venus: sync with threaded IRQ during inst destruction Sergey Senozhatsky
2024-10-24  8:59   ` Bryan O'Donoghue
2024-10-24  9:39     ` Sergey Senozhatsky
2024-10-24  9:43       ` Bryan O'Donoghue
2024-10-24 10:08         ` Sergey Senozhatsky
2024-10-24 10:52           ` Bryan O'Donoghue
2024-10-24 11:36           ` Sergey Senozhatsky [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20241024113649.GO1279924@google.com \
    --to=senozhatsky@chromium.org \
    --cc=bryan.odonoghue@linaro.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-media@vger.kernel.org \
    --cc=quic_vgarodia@quicinc.com \
    --cc=stanimir.k.varbanov@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).