summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Update to 2.6.34. Octane support is actually marked as testing in this ↵Joshua Kinard2010-06-072-0/+68
| | | | | | release (persuant to the last log entry). The patches apply, but have not been built. New patches are in the works, as time permits. Current hold-up is a lockup in the 8250 Serial Core when the system boots. This release also adds an experimental RTC Driver for SGI O2 systems and their Dallas/Maxim DS1685/DS1687 timekeeping chips. Slated to be submitted upstream soon. (Portage version: 2.1.8.3/cvs/Linux x86_64)
* Update to 2.6.27.21, 2.6.28.9, and add the newly-released 2.6.29 ebuild, ↵Joshua Kinard2009-03-241-0/+60
| | | | | | plus an updated src_unpack eblit to factor in a new local USE flag -- impactdebug. USE 'impactdebug' will apply a patch that enables a very-early, direct framebuffer device for Octane and IP28 systems that have an ImpactSR or SolidImpact graphics boardset. It is useful for debugging kernel problems that occur before the console or serial drivers are ready. (Portage version: 2.1.6.10/cvs/Linux mips64)
* Add the eblits.Joshua Kinard2009-03-1611-0/+364
| | | | (Portage version: 2.1.6.8/cvs/Linux mips64)
* Remove all old-style digests from the system and regen the Manifest files.Robin H. Johnson2008-01-312-24/+0
|
* Removed 2.6.16.* and 2.6.17.* ebuilds; Upgraded 2.6.20 to .18 from .14; ↵Joshua Kinard2007-09-035-36/+24
| | | | | | Added 2.6.22.6 to unstable for testing. Also tweaked pkg_setup to allow for marking of machines as needing 'testing' -- that is, booting on real hardware, as I can't always test each machine that I forward-port the patches for. (Portage version: 2.1.3.7)
* Update 2.6.20.13 to 2.6.20.14Joshua Kinard2007-06-141-3/+3
| | | | (Portage version: 2.1.2.9)
* Update 2.6.20.11 to 2.6.20.13.Joshua Kinard2007-06-091-3/+3
| | | | (Portage version: 2.1.2.9)
* Update the patchset with a patch to bring the uber experimental IP32 R10k ↵Joshua Kinard2007-05-201-3/+3
| | | | | | support up to par, as well as update the info text of the various machines. No rev bump needed as no sane person should be using the IP32 R10k system anyways. (Portage version: 2.1.2.7)
* Update 2.6.20.10 to 2.6.20.11Joshua Kinard2007-05-091-3/+3
| | | | (Portage version: 2.1.2.7)
* Update 2.6.20.7 to 2.6.20.10 to fix several new security issues. IP28 ↵Joshua Kinard2007-04-302-12/+12
| | | | | | support has been enabled for this revission too, however it may be flakey. Problems should be filed at Bugzilla. (Portage version: 2.1.2.5)
* Update 2.6.20.7 to a newer snapshot from linux-mips git and fix the IP27 ↵Joshua Kinard2007-04-166-39/+36
| | | | | | patches so that they actually apply. IP28 is still broke, however. Also update 2.6.17.10 -> 2.6.17.14 and 2.6.16.27 -> 2.6.16.47, and provide newer snapshots for them as well, plus a squashfs update for 2.6.16. (Portage version: 2.1.2.3)
* Update 2.6.20.6 to 2.6.20.7Joshua Kinard2007-04-141-3/+3
| | | | (Portage version: 2.1.2.3)
* So after several versions of silence, mips-sources returns to life with the ↵Joshua Kinard2007-04-071-0/+12
| | | | | | introduction of an ebuild for 2.6.20.6. This release contains an enhancement for IP22's scsi driver which may help boost speed. It also alters IP32's build methology again -- CONFIG_BUILD_ELF64 is now dead and removed from Kconfig, however the vmlinux.32 target will still be used, and a new revision of the squashfs patch has been added. The only downside is IP28 support has been disabled until new patches for it are available. Forward porting the existing patches is improbable due to changes from using CPHYSADDR to a new __pa() macro. IP27 support is untested, but assumed to work until determined otherwise (due to a lack of hardware). (Portage version: 2.1.2.3)
* Removed 2.6.15-r2, Marked 2.6.17.10 stable on mips.Joshua Kinard2006-09-051-15/+0
| | | | | (Portage version: 2.1.1_rc1-r4) (Unsigned Manifest commit)
* Update 2.6.17.7 to 2.6.17.10 (2.6.17.11 has an unresolved build-time error).Joshua Kinard2006-08-271-12/+0
| | | | | (Portage version: 2.1.1_pre5-r1) (Unsigned Manifest commit)
* Update 2.6.17.7 to 2.6.17.10 (2.6.17.11 has an unresolved build-time error).Joshua Kinard2006-08-271-0/+12
| | | | (Portage version: 2.1.1_pre5-r1)
* Update 2.6.16.22 to 2.6.16.27 and 2.6.17.3 to 2.6.17.7. This includes some ↵Joshua Kinard2006-07-262-27/+0
| | | | | | | newer patches to fix up IP27 support a little, and one or two generic patches. (Portage version: 2.1.1_pre2-r8) (Unsigned Manifest commit)
* Update 2.6.16.22 to 2.6.16.27 and 2.6.17.3 to 2.6.17.7. This includes some ↵Joshua Kinard2006-07-262-0/+27
| | | | | | newer patches to fix up IP27 support a little, and one or two generic patches. (Portage version: 2.1.1_pre2-r8)
* The 1.22 version of generic-patches tarball got damaged in getting out to ↵Joshua Kinard2006-07-061-15/+0
| | | | | | | the mirrors, so rather than wait for that to be cleared, 1.23 is being uploaded instead and all versions will reference it. Also updated 2.6.15.7 to pull in the squashfs patch and a hack to allow 64bit kernel builds on ip22. (Portage version: 2.1-r1) (Unsigned Manifest commit)
* The 1.22 version of generic-patches tarball got damaged in getting out to ↵Joshua Kinard2006-07-063-6/+21
| | | | | | the mirrors, so rather than wait for that to be cleared, 1.23 is being uploaded instead and all versions will reference it. Also updated 2.6.15.7 to pull in the squashfs patch and a hack to allow 64bit kernel builds on ip22. (Portage version: 2.1-r1)
* Update 2.6.16.20 to 2.6.16.22, import 2.6.17.3, and mask 2.6.15-r7 for all ↵Joshua Kinard2006-07-031-15/+0
| | | | | | | systems except IP22, as this is the last known kernel that worked on IP22 in 64bit mode. 2.6.16.22 and 2.6.17.3 include SquashFS for the first time, as well as being the barrier between linuxthreads and nptl support. This means 2.6.17 is enabled only in the 2006.1 nptl sub-profiles, and masked in all other profiles. It is intended to be used with a transitional linuxthreads->nptl userland, as linuxthreads support will die with 2.6.16*. (Portage version: 2.1-r1) (Unsigned Manifest commit)
* Update 2.6.16.20 to 2.6.16.22, import 2.6.17.3, and mask 2.6.15-r7 for all ↵Joshua Kinard2006-07-032-0/+27
| | | | | | systems except IP22, as this is the last known kernel that worked on IP22 in 64bit mode. 2.6.16.22 and 2.6.17.3 include SquashFS for the first time, as well as being the barrier between linuxthreads and nptl support. This means 2.6.17 is enabled only in the 2006.1 nptl sub-profiles, and masked in all other profiles. It is intended to be used with a transitional linuxthreads->nptl userland, as linuxthreads support will die with 2.6.16*. (Portage version: 2.1-r1)
* Update 2.6.16.17 to 2.6.16.20.Joshua Kinard2006-06-071-15/+0
| | | | | (Portage version: 2.1_rc4-r3) (Unsigned Manifest commit)
* Update 2.6.16.17 to 2.6.16.20.Joshua Kinard2006-06-071-0/+15
| | | | (Portage version: 2.1_rc4-r3)
* Update 2.6.16.16 to 2.6.16.17.Joshua Kinard2006-05-211-15/+0
| | | | | (Portage version: 2.1_rc1-r2) (Unsigned Manifest commit)
* Update 2.6.16.16 to 2.6.16.17.Joshua Kinard2006-05-211-0/+15
| | | | (Portage version: 2.1_rc1-r2)
* Update 2.6.16.13 to 2.6.16.16, and stable 2.6.16.*Joshua Kinard2006-05-111-15/+0
| | | | | (Portage version: 2.1_pre10-r4) (Unsigned Manifest commit)
* Update 2.6.16.13 to 2.6.16.16, and stable 2.6.16.*Joshua Kinard2006-05-111-0/+15
| | | | (Portage version: 2.1_pre10-r4)
* Update 2.6.16.11 to 2.6.16.13.Joshua Kinard2006-05-041-15/+0
| | | | | (Portage version: 2.1_pre10-r2) (Unsigned Manifest commit)
* Update 2.6.16.11 to 2.6.16.13.Joshua Kinard2006-05-041-0/+15
| | | | (Portage version: 2.1_pre10-r2)
* Fixing SHA256 digest, pass fourAlec Warner2006-04-271-1/+1
| | | | (Portage version: 2.1_pre9-r4)
* Update 2.6.16.9 to 2.6.16.11.Joshua Kinard2006-04-251-15/+0
| | | | | (Portage version: 2.1_pre9-r3) (Unsigned Manifest commit)
* Update 2.6.16.9 to 2.6.16.11.Joshua Kinard2006-04-251-0/+15
| | | | (Portage version: 2.1_pre9-r3)
* Update 2.6.16.4 to 2.6.16.9.Joshua Kinard2006-04-221-3/+3
| | | | (Portage version: 2.1_pre7-r5)
* We forgot to remove a reference to 'soft_cursor' in the Octane Odyssey video ↵Joshua Kinard2006-04-132-6/+6
| | | | | | driver, triggering a compile failure. (Portage version: 2.1_pre7-r4)
* Update 2.6.16.3 to 2.6.16.4.Joshua Kinard2006-04-121-3/+3
| | | | (Portage version: 2.1_pre7-r4)
* Update 2.6.16.2 to 2.6.16.3.Joshua Kinard2006-04-111-3/+3
| | | | (Portage version: 2.1_pre7-r4)
* Remove 2.6.14.7, and import 2.6.16.2.Joshua Kinard2006-04-112-5/+15
| | | | (Portage version: 2.1_pre7-r4)
* Removed 2.4.31 -- this terminates support for the 2.4 kernel series on all ↵Joshua Kinard2006-03-283-9/+15
| | | | | | Gentoo-supported MIPS systems. Also update 2.6.15 to the .7 release from .4. 2.6.16 import is forth coming. (Portage version: 2.1_pre7-r1)
* Another long-awaited mips-sources update. This removes 2.6.13.5, and adds ↵Joshua Kinard2006-02-143-6/+6
| | | | | | in 2.6.14.7. 2.6.15.4 is also imported with newer ebuild code to handle when a particular machine is known to fail for a given release. 2.6.15.4 also gets cobalt support working again, and includes a highly expermental patch for those interested in working on R10K-based SGI O2 systems. (Portage version: 2.1_pre3-r1)
* Another 2.6.14 stable bump, this time to .6. Patches several security flaws ↵Joshua Kinard2006-01-131-2/+2
| | | | | | from upstream, and incorporates a fix for Octane to fix a keyboard detection issue that was reported. (Portage version: 2.1_pre3-r1)
* Bump to 2.6.14.5 to plug a number of security holes, plug a compile error on ↵Joshua Kinard2006-01-031-2/+2
| | | | | | IP30 (Bug #116368), and roll in a new IP28 patchset that fixes some IMpactSR DMA issues. (Portage version: 2.0.53)
* Minor tweaks to change CVSDATE to GITDATE, and use newer rev of mips-sources ↵Joshua Kinard2005-12-201-1/+1
| | | | | | patchball due to a bug in one patch that resulted in patch failure. (Portage version: 2.0.53)
* Import 2.6.13.5 and 2.6.14.4. Removing 2.6.12.5 as IP27 now works in both ↵Joshua Kinard2005-12-203-6/+6
| | | | | | 2.6.13/2.6.14. 2.6.14.4 also brings in X support for the Impact card in IP28 (Indigo2 Impact) systems. (Portage version: 2.0.53)
* Import new kernel versions: 2.6.13.4 for ip22, ip28, ip30, and ip32;; ↵Joshua Kinard2005-10-166-12/+14
| | | | | | 2.6.12.5 for ip27 (2.6.13.4 has major PCI issues on ip27); Also updated 2.4.31 for ip22 w/ a patch that allows use of the RJ-45 port on Challenge S. (Portage version: 2.0.53_rc5)
* Added 2.6.12 final. Includes a number of enhancements: Octane (IP30) users ↵Joshua Kinard2005-07-062-5/+4
| | | | | | will discover qla1280 support and VPro support; O2 (IP32) users will notice slightly increased scsi; and Origin (IP27) users will discover sources that *gasp* actually work! (dramatic cue). (Portage version: 2.0.51.22-r1)
* New ebuilds: dropped 2.4.28 due to age and replaced it with 2.4.31 for IP22 ↵Joshua Kinard2005-06-116-16/+10
| | | | | | systems only. revbumped 2.6.10 to plug some security holes (CAN-2005-1041, CAN-2005-1265, CAN-2005-1589, etc..), and added a new ebuild for 2.6.12-rc6. (Portage version: 2.0.51.22-r1)
* Patch the 2.6 sources up a little bit. Add in a patch to finally nail that ↵Joshua Kinard2005-05-185-11/+11
| | | | | | annoying gbefb typo that kills IP32 builds, a couple security patches (CAN-2005-1263/1264/1368), and a new CVS diff for 2.6.12-rc3. (Portage version: 2.0.51.21)
* Import of an ebuild for 2.6.12-rc3Joshua Kinard2005-05-011-0/+5
| | | | (Portage version: 2.0.51.20-r4)
* The patch for IP28 was b0rked, and we had an incredibly stupid typo in the ↵Joshua Kinard2005-04-251-1/+1
| | | | | | ebuild. The kind that gets you laughed at by even novice bash coders. (Portage version: 2.0.51.19)