Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <9586bba7-ffb8-945f-eb30-88629e34737d@linux.intel.com>
Date:   Wed, 2 Mar 2022 10:37:18 +0000
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.5.0
Subject: Re: [PATCH] drm/i915/gt: Handle errors for i915_gem_object_trylock
Content-Language: en-US
To:     Jiasheng Jiang <jiasheng@iscas.ac.cn>, jani.nikula@linux.intel.com,
        joonas.lahtinen@linux.intel.com, rodrigo.vivi@intel.com,
        airlied@linux.ie, daniel@ffwll.ch, radhakrishna.sripada@intel.com,
        ville.syrjala@linux.intel.com, matthew.d.roper@intel.com,
        lucas.demarchi@intel.com, jose.souza@intel.com,
        Thomas Hellstrom <thomas.hellstrom@intel.com>,
        Matthew Auld <matthew.auld@intel.com>
Cc:     intel-gfx@lists.freedesktop.org, dri-devel@lists.freedesktop.org,
        linux-kernel@vger.kernel.org
References: <20220302061946.46630-1-jiasheng@iscas.ac.cn>
From:   Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Organization: Intel Corporation UK Plc
In-Reply-To: <20220302061946.46630-1-jiasheng@iscas.ac.cn>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk


+ Thomas, Matt

On 02/03/2022 06:19, Jiasheng Jiang wrote:
> As the potential failure of the i915_gem_object_trylock(),
> it should be better to check it and return error if fails.
> 
> Fixes: 94ce0d65076c ("drm/i915/gt: Setup a default migration context on the GT")
> Signed-off-by: Jiasheng Jiang <jiasheng@iscas.ac.cn>
> ---
>   drivers/gpu/drm/i915/gt/selftest_migrate.c | 6 +++++-
>   1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/i915/gt/selftest_migrate.c b/drivers/gpu/drm/i915/gt/selftest_migrate.c
> index fa4293d2944f..79c6c68f7316 100644
> --- a/drivers/gpu/drm/i915/gt/selftest_migrate.c
> +++ b/drivers/gpu/drm/i915/gt/selftest_migrate.c
> @@ -465,7 +465,11 @@ create_init_lmem_internal(struct intel_gt *gt, size_t sz, bool try_lmem)
>   			return obj;
>   	}
>   
> -	i915_gem_object_trylock(obj, NULL);

Guys why is this a trylock to start with? (Since being added in 
94ce0d65076c ("drm/i915/gt: Setup a default migration context on the GT").

Surely it can't ever fail since the object has just been created.

Regards,

Tvrtko

> +	if (!i915_gem_object_trylock(obj, NULL)) {
> +		i915_gem_object_put(obj);
> +		return ERR_PTR(-EBUSY);
> +	}
> +
>   	err = i915_gem_object_pin_pages(obj);
>   	if (err) {
>   		i915_gem_object_unlock(obj);