From: Lynne <dev@lynne.ee>
To: ffmpeg-devel@ffmpeg.org
Subject: Re: [FFmpeg-devel] [PATCH] aactab: add and initialize 2D VLC tables for USAC Mps212
Date: Mon, 5 May 2025 17:02:07 +0200
Message-ID: <6b235b46-0dbf-488d-931a-7a38af5c58cb@lynne.ee> (raw)
In-Reply-To: <AS8P250MB0744E64372DA0E10DE13A3918F8E2@AS8P250MB0744.EURP250.PROD.OUTLOOK.COM>
[-- Attachment #1.1.1.1: Type: text/plain, Size: 6965 bytes --]
On 05/05/2025 16:57, Andreas Rheinhardt wrote:
> Lynne:
>> On 05/05/2025 15:52, Andreas Rheinhardt wrote:
>>> Lynne:
>>>> ---
>>>> libavcodec/aac/aacdec_tab.c | 54 ++
>>>> libavcodec/aactab.c | 1820 +++++++++++++++++++++++++++++++++++
>>>> libavcodec/aactab.h | 20 +
>>>> 3 files changed, 1894 insertions(+)
>>>>
>>>
>>> 1. This should only be applied if it is used which this patch does not.
>>
>> Just posting this for reviews.
>>>> diff --git a/libavcodec/aac/aacdec_tab.c b/libavcodec/aac/aacdec_tab.c
>>>> index 45a84a9a72..5ba20c0d8a 100644
>>>> --- a/libavcodec/aac/aacdec_tab.c
>>>> +++ b/libavcodec/aac/aacdec_tab.c
>>>> @@ -111,6 +111,16 @@ const AVChannelLayout ff_aac_ch_layout[] = {
>>>> VLCElem ff_vlc_scalefactors[352];
>>>> const VLCElem *ff_vlc_spectral[11];
>>>> +const VLCElem *ff_vlc_cld_lav3_2D[2][2];
>>>> +const VLCElem *ff_vlc_cld_lav5_2D[2][2];
>>>> +const VLCElem *ff_vlc_cld_lav7_2D[2][2];
>>>> +const VLCElem *ff_vlc_cld_lav9_2D[2][2];
>>>> +
>>>> +const VLCElem *ff_vlc_icc_lav1_2D[2][2];
>>>> +const VLCElem *ff_vlc_icc_lav3_2D[2][2];
>>>> +const VLCElem *ff_vlc_icc_lav5_2D[2][2];
>>>> +const VLCElem *ff_vlc_icc_lav7_2D[2][2];
>>>> +
>>>> /// Huffman tables for SBR
>>>> static const uint8_t sbr_huffman_tab[][2] = {
>>>> @@ -279,6 +289,50 @@ static av_cold void aacdec_common_init(void)
>>>> 0);
>>>> }
>>>> +#define LAV_N_PAIR(NAME, NB) \
>>>> + ff_vlc_ ## NAME[0][0] = ff_vlc_init_tables(&state, NB, NB, \
>>>> + ff_aac_ ## NAME ##
>>>> _0_0_bits, \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _0_0_bits), \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _0_0_bits), \
>>>> + ff_aac_ ## NAME ##
>>>> _0_0_codes, \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _0_0_codes), \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _0_0_codes), \
>>>> + 0); \
>>>> + ff_vlc_ ## NAME[0][1] = ff_vlc_init_tables(&state, NB, NB, \
>>>> + ff_aac_ ## NAME ##
>>>> _0_1_bits, \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _0_1_bits), \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _0_1_bits), \
>>>> + ff_aac_ ## NAME ##
>>>> _0_1_codes, \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _0_1_codes), \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _0_1_codes), \
>>>> + 0); \
>>>> + ff_vlc_ ## NAME[1][0] = ff_vlc_init_tables(&state, NB, NB, \
>>>> + ff_aac_ ## NAME ##
>>>> _1_0_bits, \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _1_0_bits), \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _1_0_bits), \
>>>> + ff_aac_ ## NAME ##
>>>> _1_0_codes, \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _1_0_codes), \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _1_0_codes), \
>>>> + 0); \
>>>> + ff_vlc_ ## NAME[1][1] = ff_vlc_init_tables(&state, NB, NB, \
>>>> + ff_aac_ ## NAME ##
>>>> _1_1_bits, \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _1_1_bits), \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _1_1_bits), \
>>>> + ff_aac_ ## NAME ##
>>>> _1_1_codes, \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _1_1_codes), \
>>>> + sizeof(ff_aac_ ##
>>>> NAME ## _1_1_codes), \
>>>> + 0);
>>>
>>> 2. I am very much surprised that it works at all (does it?). You use the
>>> same state and therefore the same static storage as before; you add new
>>> VLCs, yet you do not increase the size of vlc_buf.
>>
>> The tables are all rather small.
>
> When I decrease the size of vlc_buf by only one element, initialization
> fails with an abort, as expected. And so it should be for you, because
> there is just no space in vlc_buf left.
>
>>
>>> 3. Can't you initialize this in a loop like ff_aac_sbr_vlc? This would
>>> remove code duplication as well as relocations.
>>
>> I didn't know that was possible. vlc.h is poorly documented, with many
>> overlapping, wrapped, renamed and macro'd functions that code from
>> different decade have gotten adapted to.
>
> Why should this not be possible? You have an example in this very
> function. And what exactly is poorly documented?
I start with the tables, not the function that initializes them. All I
knew is that VLCs are required to be in separate codes+bits tables.
There are multiple variants of VLC init, so I have no idea whether one
variant can be adapted to another.
Again, I'm just posting this for reviews.
[-- Attachment #1.1.1.2: OpenPGP public key --]
[-- Type: application/pgp-keys, Size: 637 bytes --]
[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 236 bytes --]
[-- Attachment #2: Type: text/plain, Size: 251 bytes --]
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
next prev parent reply other threads:[~2025-05-05 15:02 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-05-05 7:18 Lynne
2025-05-05 13:52 ` Andreas Rheinhardt
2025-05-05 14:48 ` Lynne
2025-05-05 14:57 ` Andreas Rheinhardt
2025-05-05 15:02 ` Lynne [this message]
2025-05-06 17:59 ` Michael Niedermayer
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=6b235b46-0dbf-488d-931a-7a38af5c58cb@lynne.ee \
--to=dev@lynne.ee \
--cc=ffmpeg-devel@ffmpeg.org \
/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
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
This inbox may be cloned and mirrored by anyone:
git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
ffmpegdev@gitmailbox.com
public-inbox-index ffmpegdev
Example config snippet for mirrors.
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git