[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <dc7d3ac8948b16f8d003828b9c4e17e7@codeaurora.org>
Date: Wed, 07 Apr 2021 11:20:16 +0530
From: dikshita@...eaurora.org
To: Bjorn Andersson <bjorn.andersson@...aro.org>
Cc: linux-media@...r.kernel.org, stanimir.varbanov@...aro.org,
linux-kernel@...r.kernel.org, linux-arm-msm@...r.kernel.org,
vgarodia@...eaurora.org
Subject: Re: [PATCH v2] media: venus : hfi: add venus image info into smem
Hi Bjorn,
Thanks for your review comments.
I have addressed all in the latest patch v3.
couldn't think of a shorter name for variables without losing the
readability so kept as it is.
Thanks,
Dikshita
On 2021-03-30 09:29, Bjorn Andersson wrote:
> On Fri 26 Mar 01:33 CDT 2021, Dikshita Agarwal wrote:
>
>> Fill fw version info into smem to be printed as part of
>> soc info.
>>
>> Signed-off-by: Dikshita Agarwal <dikshita@...eaurora.org>
>>
>> Changes since v1:
>> adressed comments from stephen.
>> removed unwanted code.
>> ---
>> drivers/media/platform/qcom/venus/hfi_msgs.c | 21
>> +++++++++++++++++++--
>> 1 file changed, 19 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/media/platform/qcom/venus/hfi_msgs.c
>> b/drivers/media/platform/qcom/venus/hfi_msgs.c
>> index 06a1908..6b6d33c9 100644
>> --- a/drivers/media/platform/qcom/venus/hfi_msgs.c
>> +++ b/drivers/media/platform/qcom/venus/hfi_msgs.c
>> @@ -6,6 +6,7 @@
>> #include <linux/hash.h>
>> #include <linux/list.h>
>> #include <linux/slab.h>
>> +#include <linux/soc/qcom/smem.h>
>> #include <media/videobuf2-v4l2.h>
>>
>> #include "core.h"
>> @@ -14,6 +15,10 @@
>> #include "hfi_msgs.h"
>> #include "hfi_parser.h"
>>
>> +#define SMEM_IMG_VER_TBL 469
>> +#define VER_STR_SZ 128
>> +#define SMEM_IMG_INDEX_VENUS 14 * 128
>
> 14 is the index, 128 is the element size, so this is now an "offset".
>
>> +
>> static void event_seq_changed(struct venus_core *core, struct
>> venus_inst *inst,
>> struct hfi_msg_event_notify_pkt *pkt)
>> {
>> @@ -239,15 +244,27 @@ static void
>> sys_get_prop_image_version(struct device *dev,
>> struct hfi_msg_sys_property_info_pkt *pkt)
>> {
>> + size_t smem_blk_sz = 0;
>
> You shouldn't need to initialize smem_blk_sz if you check the return
> value of qcom_smem_get() first.
>
>> + u8 *smem_tbl_ptr;
>> + u8 *img_ver;
>> int req_bytes;
>>
>> req_bytes = pkt->hdr.size - sizeof(*pkt);
>>
>> - if (req_bytes < 128 || !pkt->data[1] || pkt->num_properties > 1)
>> + if (req_bytes < VER_STR_SZ || !pkt->data[1] || pkt->num_properties >
>> 1)
>> /* bad packet */
>> return;
>>
>> - dev_dbg(dev, VDBGL "F/W version: %s\n", (u8 *)&pkt->data[1]);
>> + img_ver = (u8 *)&pkt->data[1];
>> +
>> + dev_dbg(dev, VDBGL "F/W version: %s\n", img_ver);
>> +
>> + smem_tbl_ptr = qcom_smem_get(QCOM_SMEM_HOST_ANY,
>> + SMEM_IMG_VER_TBL, &smem_blk_sz);
>
> 80 chars is just a guideline and this looks prettier if you avoid the
> line wrap. That said, if you pick shorter names for smem_tbl_ptr and
> smem_blk_sz you probably even have to worry.
>
>> + if ((SMEM_IMG_INDEX_VENUS + VER_STR_SZ) <= smem_blk_sz &&
>> + smem_tbl_ptr)
>
> In English you're trying to determine: "did qcom_smem_get() return a
> valid pointer and is the item's size at least as big as we need".
>
> So just write that in C:
>
> if (smem_tbl_ptr && smem_blk_sz >= SMEM_IMG_INDEX_VENUS + VER_STR_SZ)
>
>> + memcpy(smem_tbl_ptr + SMEM_IMG_INDEX_VENUS,
>> + img_ver, VER_STR_SZ);
>
> Again, please avoid the line wrap...
>
> Regards,
> Bjorn
>
>> }
>>
>> static void hfi_sys_property_info(struct venus_core *core,
>> --
>> 2.7.4
>>
Powered by blists - more mailing lists