[prev in list] [next in list] [prev in thread] [next in thread]
List: kde-panel-devel
Subject: D10747: Implement zwp_linux_dmabuf_v1
From: Roman Gilg <noreply () phabricator ! kde ! org>
Date: 2018-03-30 19:48:27
Message-ID: 20180330194827.1.3F165F7FD0D78F69 () phabricator ! kde ! org
[Download RAW message or body]
romangg added a comment.
Regarding the "drm_fourcc.h" file: do we want to copy it in KWayland's code base or \
could we use the system one? It's only available on Linux? In this case could we \
include it as a dummy only on non-Linunx systems? This way we could use the system \
one on Linux.
Should LinuxDmabufParams subclass Resource?
INLINE COMMENTS
> fredrik wrote in linuxdmabuf_v1_interface.h:39
> Do we want these nested namespaces? We could have LinuxDmabufFlags, \
> LinuxDmabufBuffer etc. instead.
Since there are not yet any namespaces in KWayland below the client/server level, I \
would prefer it without the namespace.
> fredrik wrote in linuxdmabuf_v1_interface.h:65
> Should the Buffer class use a d-pointer?
I think yes. Together with a Private class implemented in the cpp file.
> fredrik wrote in linuxdmabuf_v1_interface.h:107
> Is this the solution we want for interfacing with the compositor?
>
> My preference would be to use std::function callbacks, with setters in \
> LinuxDmabufUnstableV1Interface. Setting up the interface could then look like this: \
> m_linuxDmabuf = m_display->createLinuxDmabufInterface(m_display);
> m_linuxDmabuf->setQuerySupportedFormats([]{ return \
> Compositor::self()->scene()->supportedDrmFormats(); });
> ...
> m_linuxDmabuf->create();
>
> This can also be extended without breaking binary compatibility. But I don't think \
> we can use std::function in frameworks. There are also BIC issues when mixing \
> different STL implementations, which we may or may not care about.
I'm not sure what's the canonical way in KWayland to do this. I assume the supported \
formats and modifiers could be saved in a field of the interface's Private class on \
creation.
The buffer could be imported through a signal to the compositor and a function in the \
interface to be called by the compositor afterwards to proceed.
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D10747
To: fredrik, #kwin, #plasma, graesslin, davidedmundson, mart
Cc: romangg, plasma-devel, #frameworks, ragreen, schernikov, michaelh, ZrenBot, \
ngraham, alexeymin, lesliezhai, ali-mohamed, jensreuterberg, abetts, eliasp, sebas, \
apol, mart, hein
[Attachment #3 (unknown)]
<table><tr><td style="">romangg added a comment.
</td><a style="text-decoration: none; padding: 4px 8px; margin: 0 8px 8px; float: \
right; color: #464C5C; font-weight: bold; border-radius: 3px; background-color: \
#F7F7F9; background-image: linear-gradient(to bottom,#fff,#f1f0f1); display: \
inline-block; border: 1px solid rgba(71,87,120,.2);" \
href="https://phabricator.kde.org/D10747">View Revision</a></tr></table><br \
/><div><div><p>Regarding the "drm_fourcc.h" file: do we want to copy it in \
KWayland's code base or could we use the system one? It's only available on \
Linux? In this case could we include it as a dummy only on non-Linunx systems? This \
way we could use the system one on Linux.</p>
<p>Should LinuxDmabufParams subclass Resource?</p></div></div><br \
/><div><strong>INLINE COMMENTS</strong><div><div style="margin: 6px 0 12px 0;"><div \
style="border: 1px solid #C7CCD9; border-radius: 3px;"><div style="padding: 0; \
background: #F7F7F7; border-color: #e3e4e8; border-style: solid; border-width: 0 0 \
1px 0; margin: 0;"><div style="color: #74777d; background: #eff2f4; padding: 6px 8px; \
overflow: hidden;"><a style="float: right; text-decoration: none;" \
href="https://phabricator.kde.org/D10747#inline-51267">View Inline</a><span \
style="color: #4b4d51; font-weight: bold;">fredrik</span> wrote in <span \
style="color: #4b4d51; font-weight: bold;">linuxdmabuf_v1_interface.h:39</span></div> \
<div style="margin: 8px 0; padding: 0 12px; color: #74777D;"><p style="padding: 0; \
margin: 8px;">Do we want these nested namespaces? We could have LinuxDmabufFlags, \
LinuxDmabufBuffer etc. instead.</p></div></div> <div style="margin: 8px 0; padding: 0 \
12px;"><p style="padding: 0; margin: 8px;">Since there are not yet any namespaces in \
KWayland below the client/server level, I would prefer it without the \
namespace.</p></div></div><br /><div style="border: 1px solid #C7CCD9; border-radius: \
3px;"><div style="padding: 0; background: #F7F7F7; border-color: #e3e4e8; \
border-style: solid; border-width: 0 0 1px 0; margin: 0;"><div style="color: #74777d; \
background: #eff2f4; padding: 6px 8px; overflow: hidden;"><a style="float: right; \
text-decoration: none;" href="https://phabricator.kde.org/D10747#inline-51265">View \
Inline</a><span style="color: #4b4d51; font-weight: bold;">fredrik</span> wrote in \
<span style="color: #4b4d51; font-weight: \
bold;">linuxdmabuf_v1_interface.h:65</span></div> <div style="margin: 8px 0; padding: \
0 12px; color: #74777D;"><p style="padding: 0; margin: 8px;">Should the Buffer class \
use a d-pointer?</p></div></div> <div style="margin: 8px 0; padding: 0 12px;"><p \
style="padding: 0; margin: 8px;">I think yes. Together with a Private class \
implemented in the cpp file.</p></div></div><br /><div style="border: 1px solid \
#C7CCD9; border-radius: 3px;"><div style="padding: 0; background: #F7F7F7; \
border-color: #e3e4e8; border-style: solid; border-width: 0 0 1px 0; margin: 0;"><div \
style="color: #74777d; background: #eff2f4; padding: 6px 8px; overflow: hidden;"><a \
style="float: right; text-decoration: none;" \
href="https://phabricator.kde.org/D10747#inline-51262">View Inline</a><span \
style="color: #4b4d51; font-weight: bold;">fredrik</span> wrote in <span \
style="color: #4b4d51; font-weight: \
bold;">linuxdmabuf_v1_interface.h:107</span></div> <div style="margin: 8px 0; \
padding: 0 12px; color: #74777D;"><p style="padding: 0; margin: 8px;">Is this the \
solution we want for interfacing with the compositor?</p>
<p style="padding: 0; margin: 8px;">My preference would be to use std::function \
callbacks, with setters in LinuxDmabufUnstableV1Interface. Setting up the interface \
could then look like this:</p>
<div class="remarkup-code-block" style="margin: 12px 0;" data-code-lang="text" \
data-sigil="remarkup-code-block"><pre class="remarkup-code" style="font: 11px/15px \
"Menlo", "Consolas", "Monaco", monospace; padding: \
12px; margin: 0; background: rgba(71, 87, 120, 0.08);">m_linuxDmabuf = \
m_display->createLinuxDmabufInterface(m_display); \
m_linuxDmabuf->setQuerySupportedFormats([]{ return \
Compositor::self()->scene()->supportedDrmFormats(); });
...
m_linuxDmabuf->create();</pre></div>
<p style="padding: 0; margin: 8px;">This can also be extended without breaking binary \
compatibility. But I don't think we can use std::function in frameworks. There \
are also BIC issues when mixing different STL implementations, which we may or may \
not care about.</p></div></div> <div style="margin: 8px 0; padding: 0 12px;"><p \
style="padding: 0; margin: 8px;">I'm not sure what's the canonical way in \
KWayland to do this. I assume the supported formats and modifiers could be saved in a \
field of the interface's Private class on creation.</p>
<p style="padding: 0; margin: 8px;">The buffer could be imported through a signal to \
the compositor and a function in the interface to be called by the compositor \
afterwards to proceed.</p></div></div></div></div></div><br \
/><div><strong>REPOSITORY</strong><div><div>R127 KWayland</div></div></div><br \
/><div><strong>REVISION DETAIL</strong><div><a \
href="https://phabricator.kde.org/D10747">https://phabricator.kde.org/D10747</a></div></div><br \
/><div><strong>To: </strong>fredrik, KWin, Plasma, graesslin, davidedmundson, mart<br \
/><strong>Cc: </strong>romangg, plasma-devel, Frameworks, ragreen, schernikov, \
michaelh, ZrenBot, ngraham, alexeymin, lesliezhai, ali-mohamed, jensreuterberg, \
abetts, eliasp, sebas, apol, mart, hein<br /></div>
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic