summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Merge with Linux 2.4.0-test10.Ralf Baechle2000-11-23846-12737/+17574
|
* Fix error return value.Ralf Baechle2000-11-221-3/+6
|
* Unify the dozen pci-dma copies we had.Ralf Baechle2000-11-229-166/+15
|
* Kill printk's, no place for debug junk.Ralf Baechle2000-11-221-2/+0
|
* IOC3 fixes. Bug report from Colin Ngam.Ralf Baechle2000-11-212-4/+5
|
* CONFIG_MIPS_UNCACHED is only selectable of UP kernels. RebuildRalf Baechle2000-11-2110-10/+12
| | | | defconfig files.
* Remove ## operator that breaks kernel compile with latest gcc.Ralf Baechle2000-11-201-4/+3
|
* More un-cobalting.Ralf Baechle2000-11-182-21/+0
|
* Remove Cobalt support from the tree. It was not near compilabilityRalf Baechle2000-11-1814-2520/+0
| | | | | let alone usability. Anybody is welcome to send patches to add Cobalt support back but next time _WORKING_.
* More Atlas & Malta bits from Carsten's patch.Ralf Baechle2000-11-186-37/+84
|
* Missing from last commit.Ralf Baechle2000-11-181-5/+11
|
* New configuration option CONFIG_MIPS_UNCACHED. Not yet selectable dueRalf Baechle2000-11-183-0/+13
| | | | | to the manuals documenting ll/sc operation as undefined for uncached memory.
* Device id of the SAA9730.Ralf Baechle2000-11-181-0/+1
|
* SAA9730 Ethernet driver. Applied with hefty stilistic changes;Ralf Baechle2000-11-186-0/+1502
| | | | original patch from Carsten.
* PCI fixes for Atlas & Malta.Ralf Baechle2000-11-182-3/+74
|
* Initialization hooks for Atlas / Malta.Ralf Baechle2000-11-181-0/+12
| | | | | Start of merging a Atlas / Malta patchset. Credits for these patches should go to Carsten Langaard of MIPS.
* So, after reaching the 9-th nesting level of ifdefs in arch/mips/-Ralf Baechle2000-11-185-6/+30
| | | | | | | | | | kernel/Makefile my tastelessness detector just freaked out. The current concept of interrupt handling just doesn't scale to an ever increasing number of target platforms. Bits of new interrupt handling code are already in CVS and just to make it more clear that I won't take patches continuing this uglyness in all eternity I rename irq.c to old-irq.c which will only be compiled in when CONFIG_ROTTEN_IRQ is selected.
* Teach /proc/cpuinfo about the various variants of IP22 class machines.Ralf Baechle2000-11-184-4/+15
|
* Hopefully R3k-proof version of sysmips(MIPS_ATOMIC_SET, ...).Ralf Baechle2000-11-181-9/+10
|
* The granted member of the semaphore must be unsigned long.Ralf Baechle2000-11-172-4/+3
|
* DDB5476 default config file from Jun Sun.Ralf Baechle2000-11-091-0/+506
|
* Fix vmlinux.ecoff build target. Why the hell was this ever working?Ralf Baechle2000-11-081-0/+3
|
* Update.Ralf Baechle2000-11-089-0/+27
|
* Change time_t into unsigned int in struct stat. This fixes theRalf Baechle2000-11-081-3/+3
| | | | | newstat / lstat / fstat returning returning January 1, 1970 for all file dates.
* newstat / newlstat / newfstat have 2 not three arguments.Ralf Baechle2000-11-082-9/+8
|
* Missing Galileo EV96100 bits. Patches from Pete Popov.Ralf Baechle2000-11-062-0/+15
|
* Partial support for Galileo EV96100 evaluation board. Patches fromRalf Baechle2000-11-0633-16/+5020
| | | | Pete Popov of Monta Vista.
* Handle .kstrtab section.Ralf Baechle2000-11-062-0/+2
|
* Define RLIM_INFINITY only when building a kernel.Ralf Baechle2000-11-062-4/+4
|
* The kernel has a emulation for unaligned accesses, so don't send SIGBUSRalf Baechle2000-11-021-7/+3
| | | | on address errors.
* Call panic instead of prom_printf.Ralf Baechle2000-11-021-1/+1
|
* Add missing flags variable.Ralf Baechle2000-11-021-0/+1
|
* Update Vrc 5476 description.Ralf Baechle2000-11-021-5/+4
|
* Check for Mips Disk Volume Header before MSDOS partitions.Ralf Baechle2000-11-021-3/+3
|
* R5432 support.Ralf Baechle2000-11-013-4/+878
|
* Fix the build after the previous patch.Keith M Wesolowski2000-10-293-7/+5
|
* Fix register corruption bug. Credits for forward porting this patchRalf Baechle2000-10-273-22/+34
| | | | go to Jun Sun Industries :-)
* Fix bug in macro arguments naming.Ralf Baechle2000-10-272-8/+6
|
* Make the default ld.script a generated file to keep maintenanceRalf Baechle2000-10-274-137/+15
| | | | reasonably easy.
* 32-bit compatibility versions of recvfrom(2) and sendmsg(2). ThisRalf Baechle2000-10-272-2/+265
| | | | | fixes the bugreports about the portmapper sucking up all CPU on Origin.
* Backport the __restore_flags fixes from mips64, clean xchg().Ralf Baechle2000-10-271-92/+54
|
* Add memory clobbers to the atomic_*_return functions. I hope thisRalf Baechle2000-10-262-19/+8
| | | | fixes the file corruption / `D' state processes people observe.
* Modified two things:Nick Pollitt2000-10-261-20/+37
| | | | | 1) Added loop to provide per cpu info like other architectures 2) Added >4K output support
* Changed cpuinfo_read_proc to support >4K.Nick Pollitt2000-10-261-7/+3
|
* Delete unused variable, fix prototype definition.Kanoj Sarcar2000-10-251-1/+1
|
* Move all the definitions for non-Linux syscalls into a separate file.Ralf Baechle2000-10-254-1943/+1960
| | | | Who cares about POSIX flavoured Risc/OS syscalls etc. anyway ...
* Fix get_wchan for real. The 64-bit version also supports 32-bit psRalf Baechle2000-10-243-21/+72
| | | | | binaries. For 64-bit kernel and 32-bit the higher 32-bit of addresses have to be stripped of all addresses.
* Attempt to fix get_wchan().Ralf Baechle2000-10-241-11/+2
|
* __restore_flags had wrong constraints. The resulting registerRalf Baechle2000-10-241-66/+61
| | | | | corruption probably wasn't really problematic. Anyway, the the fixed version shaves of a few kb (!) of kernel size. Overhaul xchg().
* Fix 32-bit pread(2) / pwrite(2) syscalls.Ralf Baechle2000-10-242-2/+69
|
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-10 06:34:27 +0000