Menu
▾
▴
mactel-linux-devel
|
From: Guillaume P. <gui...@gm...> - 2007-03-17 12:52:27
Attachments:
mactel-patches-2.6.20.tgz
|
Hi, Attached to this mail you'll can find clean mactel patches for linux kernel 2.6.20. They apply against linux 2.6.20 + gentoo patchset + suspend2 without offset nor fuzz (should apply to unpatched official kernel as well). Also: - rewritten usbhid.patch which was partly applied to official 2.6.20, and added HID_QUIRK_IGNORE_MOUSE everywhere it was useful. - removed a part of applesmc.patch which was redundant with coretemp.patch I hope you will find it useful :) Regards, |
|
From: Nicolas B. <ni...@bo...> - 2007-03-17 19:03:45
|
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, Guillaume Pujol wrote: > Hi, > > Attached to this mail you'll can find clean mactel patches for linux > kernel 2.6.20. > They apply against linux 2.6.20 + gentoo patchset + suspend2 without > offset nor fuzz (should apply to unpatched official kernel as well). Ok cool. Personally, when I did this job, I edited the patchfiles manually... Then it was easier for me to do a diff between the files, and check I didn't remove a part of the patch, or modify something... With your patchset it's hard for me to compare... > Also: > - rewritten usbhid.patch which was partly applied to official 2.6.20, > and added HID_QUIRK_IGNORE_MOUSE everywhere it was useful. Just beware of this in include/linux/hid.h: +#define HID_QUIRK_IGNORE_MOUSE 0x00010000 #define HID_QUIRK_BAD_RELATIVE_KEYS 0x00010000 These two constants should have a different values. > - removed a part of applesmc.patch which was redundant with coretemp.patch Which part exactly? I cannot see what you changed because you changed all the tabs in spaces... Also I don't think there is anything in applesmc redundant with coretemp... Thanks, Best regards, Nicolas -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.3 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFF/Dt401ajQnpJXgERAhj0AJ9K0PPWyGwMQpBb5bA1PoqtGdud9ACePvvF JSd54h1L3S32qQdbws4mV/E= =2t8W -----END PGP SIGNATURE----- |
|
From: Guillaume P. <gui...@gm...> - 2007-03-17 19:42:19
|
2007/3/17, Nicolas Boichat <ni...@bo...>: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi, > > Guillaume Pujol wrote: > > Hi, > > > > Attached to this mail you'll can find clean mactel patches for linux > > kernel 2.6.20. > > They apply against linux 2.6.20 + gentoo patchset + suspend2 without > > offset nor fuzz (should apply to unpatched official kernel as well). > > Ok cool. Personally, when I did this job, I edited the patchfiles > manually... Then it was easier for me to do a diff between the files, > and check I didn't remove a part of the patch, or modify something... > With your patchset it's hard for me to compare... Actually I did almost all modifications manually directly to the patchfiles. I had an hard time tracking and fixing all tab/spaces mixings :). Apart from that and usbhid.patch, I did not change anything in the code. > > > Also: > > - rewritten usbhid.patch which was partly applied to official 2.6.20, > > and added HID_QUIRK_IGNORE_MOUSE everywhere it was useful. > > Just beware of this in include/linux/hid.h: > +#define HID_QUIRK_IGNORE_MOUSE 0x00010000 > #define HID_QUIRK_BAD_RELATIVE_KEYS 0x00010000 > > These two constants should have a different values. You are right, but this is the value figuring in the original patchfile in mactel SVN. Maybe you'd like to change this value to 0x00020000 like in upstream 2.6.21 ? > > - removed a part of applesmc.patch which was redundant with coretemp.patch > Which part exactly? I cannot see what you changed because you changed > all the tabs in spaces... Also I don't think there is anything in > applesmc redundant with coretemp... Oops sorry nervermind... I just checked again after a fresh checkout and obviously there is nothing redundant. I guess it was the result of a wrong manipulation on my local repo :| > > Thanks, Thanks to you, > > Best regards, > P.S: also I just checked and these patches apply just fine against linux-2.6.20.3 straight from kernel.org. |
|
From: Nicolas B. <ni...@bo...> - 2007-03-18 12:25:07
|
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, Guillaume Pujol wrote: > 2007/3/17, Nicolas Boichat <ni...@bo...>: >> -----BEGIN PGP SIGNED MESSAGE----- >> Hash: SHA1 >> >> Hi, >> >> Guillaume Pujol wrote: >>> Hi, >>> >>> Attached to this mail you'll can find clean mactel patches for linux >>> kernel 2.6.20. >>> They apply against linux 2.6.20 + gentoo patchset + suspend2 without >>> offset nor fuzz (should apply to unpatched official kernel as well). >> Ok cool. Personally, when I did this job, I edited the patchfiles >> manually... Then it was easier for me to do a diff between the files, >> and check I didn't remove a part of the patch, or modify something... >> With your patchset it's hard for me to compare... > > Actually I did almost all modifications manually directly to the > patchfiles. It's not really what I see when I run a simple diff between the last SVN and your patchset, there are a huge amount of differences. > I had an hard time tracking and fixing all tab/spaces > mixings :). Just don't fix anything... It makes huge patches, and you lose track of the essential modifications... It's there are spaces, ok, it's bad, but leave them there. (in the kernel, you should only have tabs) > Apart from that and usbhid.patch, I did not change > anything in the code. Ok. >>> Also: >>> - rewritten usbhid.patch which was partly applied to official 2.6.20, >>> and added HID_QUIRK_IGNORE_MOUSE everywhere it was useful. >> Just beware of this in include/linux/hid.h: >> +#define HID_QUIRK_IGNORE_MOUSE 0x00010000 >> #define HID_QUIRK_BAD_RELATIVE_KEYS 0x00010000 >> >> These two constants should have a different values. > You are right, but this is the value figuring in the original > patchfile in mactel SVN. Maybe you'd like to change this value to > 0x00020000 like in upstream 2.6.21 ? Yes. >>> - removed a part of applesmc.patch which was redundant with coretemp.patch >> Which part exactly? I cannot see what you changed because you changed >> all the tabs in spaces... Also I don't think there is anything in >> applesmc redundant with coretemp... > > Oops sorry nervermind... I just checked again after a fresh checkout > and obviously there is nothing redundant. I guess it was the result of > a wrong manipulation on my local repo :| Ok. >> Thanks, > Thanks to you, >> Best regards, >> > > P.S: also I just checked and these patches apply just fine against > linux-2.6.20.3 straight from kernel.org. Ok cool to hear it. Best regards, Nicolas -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.3 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFF/S9701ajQnpJXgERAvNDAJ9QHke3mIKRZ3rj1kPICrlKGeryZgCfXYEq l9RLNh25Kh9P4VyQHkfgSuQ= =vJLp -----END PGP SIGNATURE----- |
|
From: Soeren S. <mac...@nn...> - 2007-03-17 19:13:13
|
On Sun, 2007-03-18 at 03:03 +0800, Nicolas Boichat wrote: > Guillaume Pujol wrote: > > Hi, > > > > Attached to this mail you'll can find clean mactel patches for linux > > kernel 2.6.20. > > They apply against linux 2.6.20 + gentoo patchset + suspend2 without > > offset nor fuzz (should apply to unpatched official kernel as well). > > Ok cool. Personally, when I did this job, I edited the patchfiles > manually... Then it was easier for me to do a diff between the files, > and check I didn't remove a part of the patch, or modify something... > With your patchset it's hard for me to compare... if this patchset is against gentoo it is useful for gentoo users but not that match for mactel-svn. also I don't see the need to even patch the kernel with suspend2 ... ususpend, i.e. s2ram/s2disk/s2both work nicely enough - no ? > > Also: > > - rewritten usbhid.patch which was partly applied to official 2.6.20, > > and added HID_QUIRK_IGNORE_MOUSE everywhere it was useful. > > Just beware of this in include/linux/hid.h: > +#define HID_QUIRK_IGNORE_MOUSE 0x00010000 > #define HID_QUIRK_BAD_RELATIVE_KEYS 0x00010000 > > These two constants should have a different values. indeed IIRC the patch that is now in 2.6.21-rcX has #define HID_QUIRK_IGNORE_MOUSE 0x00020000 and while we are at it the mouse button emulation code is now also in 2.6.20.3 ... Soeren |
|
From: Nicolas B. <ni...@bo...> - 2007-03-18 12:18:05
|
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, Soeren Sonnenburg wrote: > On Sun, 2007-03-18 at 03:03 +0800, Nicolas Boichat wrote: >> Guillaume Pujol wrote: >>> Hi, >>> >>> Attached to this mail you'll can find clean mactel patches for linux >>> kernel 2.6.20. >>> They apply against linux 2.6.20 + gentoo patchset + suspend2 without >>> offset nor fuzz (should apply to unpatched official kernel as well). >> Ok cool. Personally, when I did this job, I edited the patchfiles >> manually... Then it was easier for me to do a diff between the files, >> and check I didn't remove a part of the patch, or modify something... >> With your patchset it's hard for me to compare... > > if this patchset is against gentoo it is useful for gentoo users but not > that match for mactel-svn. also I don't see the need to even patch the > kernel with suspend2 ... ususpend, i.e. s2ram/s2disk/s2both work nicely > enough - no ? Personally I use suspend2 patchset (without much success since a long time unfortunately...). Anyway, it doesn't hurt, mactel patches and suspend2 patches modify different files. >>> Also: >>> - rewritten usbhid.patch which was partly applied to official 2.6.20, >>> and added HID_QUIRK_IGNORE_MOUSE everywhere it was useful. >> Just beware of this in include/linux/hid.h: >> +#define HID_QUIRK_IGNORE_MOUSE 0x00010000 >> #define HID_QUIRK_BAD_RELATIVE_KEYS 0x00010000 >> >> These two constants should have a different values. > > indeed IIRC the patch that is now in 2.6.21-rcX has > > #define HID_QUIRK_IGNORE_MOUSE 0x00020000 > > and while we are at it the mouse button emulation code is now also in > 2.6.20.3 ... Ok. Best regards, Nicolas -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.3 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFF/S3M01ajQnpJXgERAneJAJ9tYMaZbHqQItgmZDzsuSVWxEFIMwCcCqqB MP9zIhEbPf5RVTyxIpcHHx8= =QPM4 -----END PGP SIGNATURE----- |
Thanks for helping keep SourceForge clean.
X