| Summary: |
Missing lines in xterms |
| Product: |
xorg
|
Reporter: |
Kamil Iskra <kamil.42920> |
| Component: |
Driver/intel | Assignee: |
Chris Wilson <chris> |
| Status: |
RESOLVED
DUPLICATE
|
QA Contact: |
Intel GFX Bugs mailing list <intel-gfx-bugs> |
| Severity: |
major
|
|
|
| Priority: |
medium
|
|
|
| Version: |
unspecified | |
|
| Hardware: |
x86-64 (AMD64) | |
|
| OS: |
Linux (All) | |
|
| Whiteboard: |
|
|
i915 platform:
|
|
i915 features:
|
|
|---|
| Attachments: |
|
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 130475 [details] Corruption visible in the bottom-left corner I'm seeing serious drawing issues in xterms (the good old fashioned ones, "xterm" binary). Screen doesn't get refreshed or lines are missing (empty). Switching to a different window fixes them. Reproducible: I don't have a foolproof testcase, but just a few minutes in a terminal window tends to be enough to notice problems. Seems to be especially common with tall (full screen height) terminal windows and when multiple lines need to drawn quickly in succession. Simply repeating 'ls -l' a dozen times or so tends to be enough, or a "cat" of a short text file. Workaround: switching to UXA or the modesetting driver. Environment: Sony Vaio Pro 11 laptop with i5-4200U CPU (HD Graphics 4400). 4.10.6 kernel, KDE plasma 5.8.5 (compositing enabled). Xorg server 1.18.4 (1.19.2 acts the same). The critical component seems to be the Intel driver. It works fine with what Gentoo calls version "2.99.917_p20160621" (based on git commit c28e62f94f15c9f5c4fb0744588f08ae18e4a9b5). The next Gentoo version is "2.99.917_p20161206" (based on git commit 169c74fa6c2cd9c28dd7bfacd9639cd245b8c8a8) and that one is already broken. I tried every subsequent Gentoo version as well as your git HEAD (cb6ba2da056f3298a765b4da5cd626343c00a533) and they all exhibit this bug. I'll attach a screenshot showing the corruption as well as my Xorg log and a dmesg output. I could try bisecting it if needed but considering a nearly six-month window, I would appreciate a list of possible culprits.