summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Merge with 2.1.56 as first part of merging back my code.Ralf Baechle1997-12-01135-2543/+4295
|
* Remove superfluous disable_irq() statement. It was causingRalf Baechle1997-11-301-1/+0
| | | | interrupts to be disabled for a very long time.
* Fixed the userland semaphores. No need for the dentry hacksMiguel de Icaza1997-11-011-150/+57
| | | | as I found that struct file has a private_data pointer. WOOO!
* bad miguel, bad bad bad miguel.Miguel de Icaza1997-09-231-1/+1
| | | | | | slap slap slap. actually return the value of vmalloc.
* init the irix inventory.Miguel de Icaza1997-09-211-0/+1
|
* More bits of the shmiq code implementation.Miguel de Icaza1997-09-211-3/+31
|
* More ioctls, more structures, more more more.Miguel de Icaza1997-09-212-0/+25
|
* Miguel goes cosmetic.Miguel de Icaza1997-09-211-1/+1
|
* Fixes to the shmiq implementation.Miguel de Icaza1997-09-212-19/+167
| | | | Fixes to some of the stream-like devices; implemented the mouse interface.
* Some newport specific ioctls implemented.Miguel de Icaza1997-09-212-3/+170
| | | | More DCB/xmap9 defines compatible with the SGI defines added;
* More toys: per-card ioctl and per-card reset console.Miguel de Icaza1997-09-211-0/+2
|
* 1. Add support for enabling/disabling console output (requiered when theMiguel de Icaza1997-09-211-27/+54
| | | | | | | | | | | | | | | X server is running. This machine does not have a crappy fb you can just play with like the sparcs). 2. Drop support for multiple cards (temporary, I need to fix the way the direct graphics code works). 3. Lots of fixes to the GFX_* ioctls. 4. Add support for per-graphics card ioctl processing. 5. Direct rendering works properly now (context switching code disabled for now, this depends on the rrm code updates I have to do).
* newport_reset() implemented. upadte to new graphics.h interface.Miguel de Icaza1997-09-211-17/+24
|
* 1. inventory support.Miguel de Icaza1997-09-211-21/+105
| | | | | | 2. Autogrow mmap flag for irix binaries bad taste hack. 3. fixed irix stat32 4. IRIX_ALLOC_SP fcntl ignored.
* Pass the ioctls down to the Linux kernel when the irix_ioctl does notMiguel de Icaza1997-09-211-1/+5
| | | | know about it.
* Map the PRDA page on IRIX processesMiguel de Icaza1997-09-211-0/+35
|
* include irixinv in the compilationMiguel de Icaza1997-09-211-1/+2
|
* The IRIX prctl interfaceMiguel de Icaza1997-09-211-0/+42
|
* IRIX inventory support. Right now it has hardcoded my machine'sMiguel de Icaza1997-09-212-0/+106
| | | | | | information. This needs to be fixed at least to report the proper graphics capabilities on the machine. Optimally, we should scatter the right calls to the inventory all over the kernel source.
* Support for the Nevada. Cleanup.Ralf Baechle1997-09-171-38/+6
|
* Fix for little endian MIPS boxes.Ralf Baechle1997-09-171-1/+1
|
* Implement microsecond timers based on the cycle counter.Ralf Baechle1997-09-171-8/+155
|
* Don't include <linux/config.h>.Ralf Baechle1997-09-171-2/+1
|
* namei.c: - Support symlinks correctly.Mike Shaver1997-09-168-160/+416
| | | | | | | | inode.c: - Formatting, minor change to start supporting indirect extents. dir.c: - Prevent access to an indirect directory from hosing the system. *.h: - Add forgotten include files.
* Added a few #ifdef CONFIG_REMOTE_DEBUG's to keep Mark's cool debuggingMike Shaver1997-09-161-1/+4
| | | | | stuff from breaking kernel compiles for those of us who aren't cool enough to have kgdb set up yet. =)
* EFS integration (read-only, slight problem with symlinks).Mike Shaver1997-09-1611-11/+1435
|
* Call set_async_breakpoint() if kgdb 'stop' character is received. kgdbMark Salter1997-09-161-2/+2
| | | | | will subsequently take control at the insn interrupted by the incoming stop character.
* Added set_async_breakpoint(). This can be called from serialMark Salter1997-09-161-4/+31
| | | | | drivers to support asynchronous interruption of kernel during kgdb sessions.
* Fixed bus error by avoiding HPC bug affecting reads of dma registers.Mark Salter1997-09-161-14/+15
| | | | Thanks for the docs, Ariel!
* Make irix_*getdents* track the readdir signature changes.Mike Shaver1997-09-131-4/+4
|
* - Set caching mode for KSEG0 to cached-noncoherent for all machines.Ralf Baechle1997-09-124-43/+92
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On some MIPS boxes the firmware doesn't do that for us. - We still had two unprotected loads in the sys_sigrestore(2). Use __get_user(). - Handle QED-style L1 caches != 16kb per cache correctly. - Protect the cacheflush instructions for the signal trampoline just like the loads in __get_user(). Otherwise the following code will result in a nice "Can not handle kernel paging request" message: #include <signal.h> static void hurz(void) { } main() { signal(SIGSEGV, hurz); /* Chainsaw the stack pointer ... */ asm("move $29,%0" : :"r"(0x70000000)); *(int *) 8 = 0; } With the fix applied I still get the error message. The cause of this problem is that gas produces an __ex_table section with wrong contents. Oh well, how good that I have a nice working post 2.8.1 binutils version at home in good ol' Germany ...
* Cleanup.Ralf Baechle1997-09-121-18/+1
|
* Remove files that were removed between 2.1.49 and 2.1.55. SomehowRalf Baechle1997-09-1239-1132/+0
| | | | my script to do that broke :-(
* Merge with Linux 2.1.55. More bugfixes and goodies from my privateRalf Baechle1997-09-12673-20551/+77319
| | | | CVS archive.
* Removed the autodestruct prevention mechanisms, now that we'reMike Shaver1997-09-112-9/+2
| | | | grown-up enough to not cause HW catastrophe.
* Mark Salter's trailing-space fix.Mike Shaver1997-09-041-1/+3
| | | | I can now boot from local disk!
* Fixed unaligned copiesMark Salter1997-09-021-5/+4
|
* Fix the ELF interpreter loader. dynamic IRIX binaries now workMiguel de Icaza1997-09-021-36/+39
| | | | again.
* Rename cp0_reg4 to cp0_contextMark Salter1997-09-011-3/+3
|
* Fixed single-stepping.Mark Salter1997-09-011-4/+123
|
* Disable interrupts for gdb stub. Also, save cp0_contextMark Salter1997-09-011-1/+6
|
* Fixed unaligned access macros. gcc's builtin memcpy won't work.Mark Salter1997-09-011-9/+47
|
* Newport context switching implemented (la de da, la de da).Miguel de Icaza1997-08-285-142/+382
| | | | | | | Linux newport structure is now compatible with the IRIX version. GNOME compliant display properties implemented. Oops. Wrong mailing list.
* tty video type numberMiguel de Icaza1997-08-261-1/+1
|
* Disable the console driver when a direct rendering applicationMiguel de Icaza1997-08-262-2/+7
| | | | is running.
* restore_flags fix from Ralf.Miguel de Icaza1997-08-261-1/+3
|
* irq_enter/irq_enter fix, as described by RalfMiguel de Icaza1997-08-261-2/+5
|
* fix for the irix elf loaderMiguel de Icaza1997-08-261-3/+3
|
* Corrected waste of miscdevice for usema (thanks, Miguel).Mike Shaver1997-08-221-2/+1
|
* Mark Salter's patch to fix the excessive eth0 timeout bug.Mike Shaver1997-08-201-22/+26
| | | | | | | | | | | | | | | | | Mark's description of the problem: This patch seems to have fixed the excessive ethernet timeouts I've been seeing. The problem was a race condition when queueing another packet for transmisson. The code in sgiseeq_start_xmit() would try to start another transmission from the head of the transmit queue if the HPC was inactive. Because interrupts are off at this time, it was possible that that packet had been transmitted but the queue head had not been advanced by an interrupt service. This resulted in a duplicate packet being sent over the wire and when interrupts were reenabled, the normal ack'ing of the original transmission of that packet would put the HPC into a wedged state which eventually led to a timeout and reset.
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-10 07:52:18 +0000