April 2010 - Wine-Devel - list.winehq.org

Re: New SCSI Driver for Mac OS X

by Charles Davis

On 4/28/10 9:35 AM, Charles Davis wrote: > I don't have a public repository yet. My plan is to create a SourceForge > project for this (like AJ suggested). And here it is: http://sourceforge.net/projects/xnu-extras/ Chip

April 28, 2010

Re: New SCSI Driver for Mac OS X

by Charles Davis

On 4/28/10 2:14 PM, Stefan Dösinger wrote: > > Am 28.04.2010 um 17:35 schrieb Charles Davis: > >> On 4/28/10 4:03 AM, Stefan Dösinger wrote: >>> >>> Am 27.04.2010 um 22:10 schrieb Charles Davis: >>> >>>> tar jxf ExtMediaBSDClient.tar.bz2 >>>> cd ExtMediaBSDClient >>> Do you have the sourcecode somewhere with instructions on how to compile it? >> Sure. I've attached the source here. To compile it, you need Xcode >> installed. > Looks interesting. I prefer to have the sourcecode before I load something into my kernel :-) That's understandable. > >> Once it's loaded, all SCSI disks that are inserted afterward will accept >> the IOCTL. > I assume that means I have to remove my cdrom from my drive and reinsert it after loading the driver, correct? Or does it mean removing the drive(as in unplugging and replugging a USB drive for example?) If it's a removable disk drive, you don't have to physically unplug it. :) Reinserting the CD-ROM should work. Chip

April 28, 2010

Re: New SCSI Driver for Mac OS X

by Stefan Dösinger

Am 28.04.2010 um 17:35 schrieb Charles Davis: > On 4/28/10 4:03 AM, Stefan Dösinger wrote: >> >> Am 27.04.2010 um 22:10 schrieb Charles Davis: >> >>> tar jxf ExtMediaBSDClient.tar.bz2 >>> cd ExtMediaBSDClient >> Do you have the sourcecode somewhere with instructions on how to compile it? > Sure. I've attached the source here. To compile it, you need Xcode > installed. Looks interesting. I prefer to have the sourcecode before I load something into my kernel :-) > Once it's loaded, all SCSI disks that are inserted afterward will accept > the IOCTL. I assume that means I have to remove my cdrom from my drive and reinsert it after loading the driver, correct? Or does it mean removing the drive(as in unplugging and replugging a USB drive for example?)

April 28, 2010

Re: shell32: add configure option --disable-unixfs

by Alexandre Julliard

Dan Kegel <dank(a)kegel.com> writes: > Fixes http://bugs.winehq.org/show_bug.cgi?id=22450 There's no reason to do that in configure. -- Alexandre Julliard julliard(a)winehq.org

April 28, 2010

Re: X11 keyboard detection bitrot on Linux?

by GOUJON Alexandre

Hi, It may not be related but I filled a similar bug at http://bugs.winehq.org/show_bug.cgi?id=21307 I sent a patch (http://source.winehq.org/patches/data/60888) wich still waits for comments. And if it was working before, as usual, git-bisect would help. Thanks

April 28, 2010

Re: [PATCH 3/4] wined3d: fix texm3x3spec coordinate mask bug

by Henri Verbeet

On 28 April 2010 19:40, Roderick Colenbrander <thunderbird2k(a)gmail.com> wrote: > I didn't make up out of that discussion that another solution was > needed but no problem. Yesterday I mentioned that 'setting fake 2d > sampler' line but after a more thorough look > pixel_shader_update_samplers which is called before the shader is > generated again sets a 2D sampler (it gets this from the texture). > This makes me believe the patch is fine unless the call which > retrieves this info from the texture is wrong but I doubt it is wrong > but if you want I can look closer to that. > Maybe it just needs a test then.

April 28, 2010

Re: [PATCH 3/4] wined3d: fix texm3x3spec coordinate mask bug

by Roderick Colenbrander

I didn't make up out of that discussion that another solution was needed but no problem. Yesterday I mentioned that 'setting fake 2d sampler' line but after a more thorough look pixel_shader_update_samplers which is called before the shader is generated again sets a 2D sampler (it gets this from the texture). This makes me believe the patch is fine unless the call which retrieves this info from the texture is wrong but I doubt it is wrong but if you want I can look closer to that. Roderick On Wed, Apr 28, 2010 at 2:54 PM, Henri Verbeet <hverbeet(a)gmail.com> wrote: > On 28 April 2010 14:29, Roderick Colenbrander <thunderbird2k(a)gmail.com> wrote: >> This patch fixes an issue in "King's Bounty: The Legend" for which a shader of the form: >> 'texture2D(sampler2D, tmp0.xyz)' was generated. The nvidia GLSL compiler wants to have tmp0.xyz. > > We talked about this on IRC yesterday, the reason this happens is that > the sampler type is wrong. > > >

April 28, 2010

Re: X11 keyboard detection bitrot on Linux?

by André Hentschel

Am 28.04.2010 16:07, schrieb Joerg-Cyril.Hoehle(a)t-systems.com: > Hi, > > Perhaps there's some bitrot that affects the handling of alt/altGr > keys on Linux? > > Using WINEDEBUG=+keyboard on Macos yields exepected results. The 3rd > and 4th characters (|\ etc.) below show what's obtained when pressing > the alt key: > trace:key:X11DRV_KEYBOARD_DetectLayout mismatch for keycode 34, got 7/|\ > trace:key:X11DRV_KEYBOARD_DetectLayout mismatch for keycode 36, got 8({ > trace:key:X11DRV_KEYBOARD_DetectLayout mismatch for keycode 58, got <>?? > trace:key:X11DRV_KEYBOARD_DetectLayout mismatch for keycode 73, got ,,.. > > On Linux instead: > warn:keyboard:X11DRV_KEYBOARD_DetectLayout 7 keysyms per keycode not supported, set to 4 > trace:key:X11DRV_KEYBOARD_DetectLayout mismatch for keycode 16, got 7/7/ > trace:key:X11DRV_KEYBOARD_DetectLayout mismatch for keycode 17, got 8(8( > > None of the {[]}\ keys reached via AltGr-7/8/9/0 on the German > keyboard is printed by the keyboard detection code. This causes > warnings for all keyboards (not printed by default) because none matches, > which presumably was not the case back when that code was written. > What has changed? > > Regards, > Jörg Höhle I cannot confirm that on 64-bit Ubuntu 10.04 Kernel 2.6.32-21-generic with wine from yesterday. -- Best Regards, André Hentschel

April 28, 2010

Re: kernel32: Add tests for lstrcmpW (try 3)

by Dmitry Timoshkov

André Hentschel <nerv(a)dawncrow.de> wrote: > + for (r = 0; r < 0xFFFF; r++) > + { > + if (ignorewchar(r)) > + { > + markerW[2] = r; > + todo_wine > + { > + ok(!lstrcmpW(markerW, nomarkerW), "expected the character 0x%x to be ignored within the string\n", r); > + ok(CompareStringW(LOCALE_SYSTEM_DEFAULT, 0, markerW, -1, nomarkerW, -1) == 2, > + "expected the character 0x%x to be ignored within the string\n", r); > + ok(CompareStringW(LOCALE_SYSTEM_DEFAULT, NORM_IGNORECASE|NORM_IGNORENONSPACE|NORM_IGNORESYMBOLS| > + SORT_STRINGSORT|NORM_IGNOREKANATYPE|NORM_IGNOREWIDTH|LOCALE_USE_CP_ACP, > + markerW, -1, nomarkerW, -1) == 2, > + "expected the character 0x%x to be ignored within the string\n", r); > + } > + } > + } This is not better, my comments for your previous attempts still apply. In addition to those: 1. if you decided to use some specific flags in the CompareString call, you may want to test other variants of them. Using LOCALE_USE_CP_ACP in a W version of the API doesn't look useful. 2. return value of CompareString has symblic names. -- Dmitry.

April 28, 2010

Re: New SCSI Driver for Mac OS X

by Charles Davis

On 4/28/10 4:03 AM, Stefan Dösinger wrote: > > Am 27.04.2010 um 22:10 schrieb Charles Davis: > >> tar jxf ExtMediaBSDClient.tar.bz2 >> cd ExtMediaBSDClient > Do you have the sourcecode somewhere with instructions on how to compile it? Sure. I've attached the source here. To compile it, you need Xcode installed. After unpacking the tarball, from the XNUKernelExtras directory, just type: xcodebuild at a Terminal. Then to install it, do: sudo xcodebuild install DSTROOT=/ There's a header that goes along with it. To install it, you do: sudo xcodebuild -target headers install DSTROOT=/ I've also attached the source for the test program. You can just do: gcc -o ExtMediaTest ExtMediaTest.c on it after installing the header above. I don't have a public repository yet. My plan is to create a SourceForge project for this (like AJ suggested). Chip

April 28, 2010