#wine-staging

ah i forgot one thing. it might be a good idea to add mlockall() to the patch too, as if you are using audio you wouldn't want anything to be swapped out, the same applies for the wineserver code

but that one could land you in a swap storm, probably more annoying :)

JackWinter: that should probably a separate patch then

i'll think about how to add it. another env var? then all of wine and the app would be locked into ram?

JackWinter: it would only make sense for the wineserver, not for the app itself

i think i concluded last time that you couldn't use the reg from the server code? might be that i was just lazy though :)

sure it makes sense for the app too

though i suppose you could add 2 vars for the server and the app?

if you run chromium and it leaks ram, then for instance you might wait less time to switch back to the wine app if it hasn't been swapped out, and like i said audio programs

i assume you would pass MCL_FUTURE, right? the problem is that its not just a recommendation for the kernel, but instead:

"""If MCL_FUTURE has been specified, then a later system call (e.g., mmap(2), sbrk(2), malloc(3)), may fail if it would cause the number of locked bytes to exceed the permitted maximum (see below)."""

and just locking one time at the beginning would basically only lock ntdll/kernel, but not the rest of the program

but if wine will ask for too much ram it might make all the rest of the system swap out. X, kwin, plasma, terminals... a swap storm is a pain, it can take like 5 minutes to get the system back...

for the wineserver it makes more sense because no further modules are loaded at runtime

yeah, in realtime terms it's good that it preallocates

but ok, let's leave that for a different patch. no need to muddy up waters :)

but on the other hand i think if the user specifically enables something, he might have to live with the consequences.... it's like that with many things in linux and computers in general :)

you can't nanny sit the users and make things perfect unless you are apple or so..

JackWinter: hehe ;)

ah, and yes MCL_FUTURE is the only way i know to propagate ram mlock, and tough luck if your progs allocate more ram than you have. then don't enable it or buy more ram.

JackWinter: it might also be killed before the memory limit is reached, afaik it can be configured how much memory an app can lock max

10sebastian@fds-team.de changed the Component on bug 183 from Discussion to Patch Submission.

04Bug http://bugs.wine-staging.com/show_bug.cgi?id=183 normal, ---, ---, bugs, UNCONFIRMED , wineserver: Add support for windows priority levels by using realtime scheduling policies

yeah, that's a PAM limit. but i have mine set to unlimited :)

but again i've never ran into that problem, just swap storms due to chromiums ram use doubling or trippling after a while, and starting up an audio project when there isn't enough ram left

i did find a program calle memlockd, which has a list of programs that it memlocks and preloads. like that X, kwin, plasma, teminal and other progs can be locked into ram and when the sh*t hits the fan, you can still quickly get on top of things, instead of waiting for 10mins and checking out what OOM chose to kill this time

do you guys have a wayland backend or plan on having one?

diegoviola: we don't have one yet, but sooner or later it will probably be necessary

ok thanks

it has more than 32.000 lines of code

this would be a huge task and moreover, i am also using the proprietary nvidia driver as default, so no wayland for me ;-)

DarkPlayer: any opinion about x-server/wayland?

how it impacts wine

wonder if anyones working on a wayland driver

jarkko: it would not offer any additional functionalities, its just a lot of work ^^ upstream wine will most likely not start unless more distros switch to wayland by default

ariscop: unlikely, but codeweavers is working on some android driver

wayland is coming android?

no, android will be a separate driver

has anyone even tried it ?

i havent

i but i constantly get some xwayland file when updating mint files via upgrades

xwayland seems to handle wine ok

mostly ok

phoronix havent mentioned wayland for a long time

i had forgotten the hole project

i am not following it somehow, so no idea what's going on

there's a really odd bug where wine windows stop accepting input, and xwaland segfaults when weston exits

does it stop getting input if you alt tab away from the program?

jarkko, weston was running inside x, couldn't alt+tab

DarkPlayer: does wine include own xserver or does it use the one users has

jarkko: this wouldn't make sense, you can not run the original x server inside of another xserver

i think crossover does something like that

x can nest but there'd be little point

i know that there are programs like Xephyr

xpra is one of the cooler examples

jarkko, crossover may have used to use their own x server on osx, but fairly sure theres a native driver now

any idea if i want to see missing dx10/11 functions of wine, where to search? msdn? or some other website?

jarkko: directx is a COM based API, but you should be able to find all interfaces on the msdn website

i am looking msdn currently

the dx10 createdevice code could be actually work

- be

lol those permission patches causes regression bug already

04Bug 14896: normal, P2, ---, wine-bugs, NEW , GTA: San Andreas - Ambient Sounds are too quiet (EAX)

eax

[06winetricks/master][057a28448](00http://is.gd/59pujs) 02Sam Reed: Update takao-fonts download url

[06winetricks/master][0583f8dcf](00http://is.gd/gG3Fg4) 02austin987: Merge pull request #490 from reedy/patch-1

^^ oo good, the bot reporting of winetricks works

[06winetricks/master][058173eb3](00http://is.gd/q0l6gF) 02Austin English: irfanview: update URL to a working one

[06winetricks/master][05a7d25c1](00http://is.gd/vvlcbg) 02Austin English: osmos_demo: demo is no longer online, remove from winetricks

10New Wine Staging bug 184 filed by austinenglish@gmail.com.

04Bug http://bugs.wine-staging.com/show_bug.cgi?id=184 normal, ---, ---, bugs, UNCONFIRMED , Xvid installer requires mfplat.dll.MFTRegister

10New Wine Staging bug 185 filed by sebastian@fds-team.de.

04Bug http://bugs.wine-staging.com/show_bug.cgi?id=185 normal, ---, ---, sebastian, CONFIRMED , EM_SETPASSWORDCHAR ineffective when edit control does not have WS_BORDER set