| Summary: | XPutImage performance regression with xlib/xcb | ||||||
|---|---|---|---|---|---|---|---|
| Product: | xorg | Reporter: | Clemens Eisserer <linuxhippy> | ||||
| Component: | Lib/Xlib | Assignee: | xcb mailing list dummy <xcb> | ||||
| Status: | RESOLVED FIXED | QA Contact: | Xorg Project Team <xorg-team> | ||||
| Severity: | normal | ||||||
| Priority: | medium | ||||||
| Version: | git | ||||||
| Hardware: | Other | ||||||
| OS: | All | ||||||
| Whiteboard: | 2011BRB_Reviewed | ||||||
| i915 platform: | i915 features: | ||||||
| Attachments: |
|
||||||
as far as I can remember, the difference was mostly caused by the different buffer sizes. I'll test it soon so see if there's still a difference... (In reply to comment #1) > as far as I can remember, the difference was mostly caused by the different > buffer sizes. I'll test it soon so see if there's still a difference... > any news? re-tested and there is no regression compared to "classic" xlib. So this bug is fixed - Thanks :) Closing as fixed. |
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.
Created attachment 19129 [details] test-case All the test-case attached does is uploading 32x32x8 bit images (1kb) and do a composite operation with the data used as mask. I was testing the advantage of using SHM and using SHM is over 2x faster using the XCB based Xlib than XPutImage. Here are my findings: Xlib/Put: 80ms Xlib/ShmPut: 75ms Xcb-Xlib/Put: 180ms (!) Xcb-Xlib/ShmPut 80ms I was using the xlib-binaries distributed by Fedora, Xlib was Fedora7, and Xcb-Xlib was on Fedora9, so please take the results with a grain of salt. I also don't know wether Fedora maybe hand-tuned the Xlib-Buffer size in Fedora7.