On Monday 15 December 2008, Jeffrey Baker wrote: > I seriously doubt it, since performance was far better last week, last > month, and a year ago with the same hardware. I think we can safely > pin this current problem on the software. > > Even if it were limited at 50kglyphs/s, the 7m character output would > still come out in 140s, not 400s. > > -jwb
I must say I'm also very unhappy with the intel driver's degrading performance. I don't have any hard numbers, but the "snappiness" of esp. scrolling text has gone nothing but downward over the last months. It's nice to see/read developers experimenting with new acceleration architectures and all, but if there is no net gain over XAA, what is it worth anyway? I certainly would not mine one of my host CPU's four cores at 50% load if my everdays' desktop tasks performance would be better than what it is. I don't really want to believe it's the hardware that's too slow for what X is doing; last I checked, Windows GDI performance was top notch with Intel hardware. I really love how Intel and its fine development team supports a free software stack even on the desktop side of things, and how these efforts benefit even users of hardware from different manufacturers. Yet on the other hand, I'd really appreciate more accurate/up-to-date docs about the good stuff that's in the making. It sucks to iterate over some half a dozen developer's blogs every second month or so just to grok why EXA is so much slower than XAA if $option is (not) set. There should be some kind of "tweaking guide" on intellinuxgraphics.org imho or at least some other definite resource with infos about what's recommended, and what's not. I'm tired of fiddling with my xorg.conf anew for every minor release that gets pushed out, just to have my system revert to 70-90% of the last driver revision's performance level. Sorry for ranting, esp. at this length... I hope you don't take it the wrong way. I really appreciate all your work, I just wish it would be more visible for us, the "end users". (I've got an GMA X3100 in my desktop, and it's seemingly growing ever slower for months now, as well as a GMA X4500HD in my ThinkPad x200 which hardlocks the system when resuming from disk with the "intel"-module loaded, so please bear with me. Btw, don't you think a fix of the latter problem, which I do think exists, would warrant for a minor bugfix release? Please? Pretty-please? :)) -- with best regards: - Johannes Truschnigg ( [email protected] ) www: http://johannes.truschnigg.info/ phone: +43 650 2 133337 jabber: [email protected] Please do not bother me with HTML-eMail or attachments. Thank you.
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ xorg mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/xorg
