- 11 Mar, 2004 2 commits
-
-
Jens Axboe authored
This patch allows you to map a request with user data for io, similarly to what you can do with bio_map_user() already to a bio. However, this goes one step further and populates the request so the user only has to fill in the cdb (almost) and put it on the queue for execution. Patch converts sg_io() to use it, next patch I'll send adapts cdrom layer to use it for zero copy cdda dma extraction.
-
Stephen Rothwell authored
While playing with udev, I discovered that the virtual console devices on iSeries had there minor numbers off by one i.e. /dev/tty1 was minor 2! This fixes it.
-
- 10 Mar, 2004 14 commits
-
-
Andi Kleen authored
From Takashi Iwai kobj_lookup had a 64bit bug, which caused the request of a unknown character device to burn CPU instead of failing quickly.
-
Andi Kleen authored
The biggest new feature is fixed 32bit vsyscall (SYSCALL+SYSENTER) support, mostly from Jakub Jelinek. This increases 32bit syscall performance greatly (latency halved and better). The SYSENTER for Intel support required some infrastructure changes, but seems to work now too. The 64bit vsyscall vtime() just references xtime.tv_sec now. This should make it a lot faster too. A fix for some Intel IA32e systems. Also a few long standing bugs in NMI like exception handlers were fixed. And a lot of other bug fixes. Full changeLog: - Clean up 32bit address room limit handling, fix 3gb personality - Move memcpy_{from,to}io export to ksyms.c file. This seems to work around a toolchain bug (Andreas Gruenbacher) - Update defconfig - ACPI merges from i386 (SBF should work now, acpi=strict) - Implement mmconfig support based on i386 code (untested) - Fix i386/x86-64 pci source file sharing - Implement ptrace access for 32bit vsyscall page - Always initialize all 32bit SYSENTER/SYSCALL MSRs. - Export run time cache line size to generic kernel - Remove explicit CPUID in ia32 syscall code - Fill in most of boot_cpu_data early - Remove unused PER_LINUX32 setup - Fix syscall trace in fast 32bit calls (Suresh B. Siddha) - Tighten first line of the oops again. - Set up ptrace registers correctly for debug,ss,double fault exceptions - Fix 64bit bug in sys_time64 - Optimize time syscall/vsyscall to only read xtime - Fix csum_partial_copy_nocheck - Remove last traces of FPU emulation - Check properly for rescheduling in exceptions with own stack - Harden exception stack entries (#SS,#NMI,#MC,#DF,#DB) against bogus GS - Use an exception stack for machine checks - Handle TIF_SINGLESTEP properly in kernel exit - Add exception stack for debug handler - Disable X86_HT for Opteron optimized builds because it pulls in ACPI_BOOT - Fix CONFIG_ACPI_BOOT compilation without CONFIG_ACPI - Fix eflags handling in SYSENTER path (Jakub Jelinek) - Use atomic counter for enable/disable_hlt - Support 32bit SYSENTER vsyscall too (Jakub Jelinek) - Don't redefine Dprintk - Change some cpu/apic id arrays to char - Support arbitary cpu<->apicid in hard_smp_processor_id (Surresh B Sidda) - Move K8 erratum #100 workaround into slow path of page fault handler. - Fix 32bit cdrom direct access ioctls (Jens Axboe) - Enable 32bit vsyscalls by default - Fix 32bit vsyscalls (Jakub Jelinek)
-
Benjamin Herrenschmidt authored
This makes the temperature control code more robust, putting less pressure on i2c, and work around occasional misconfiguration of the ADC chips leading to incorrect temperature readings.
-
Linus Torvalds authored
-
Benjamin Herrenschmidt authored
This patch fixes the code that workaround lockups when accessing the config space of devices on K2 when they are shut down. The code was there but in the wrong place ;) And a typo prevented the ohci1394 version of it from working.
-
bk://gkernel.bkbits.net/net-drivers-2.5Linus Torvalds authored
into ppc970.osdl.org:/home/torvalds/v2.5/linux
-
Don Fry authored
When adding an additional device, a variable which has been removed is accessed. Adding the first device or a 4-port card works. Adding a second device was not done in PCI hot plug testing. Now tested on IA32.
-
Linus Torvalds authored
The bug turned out to be memory corruption by DMA, and the corrupted address made it that much more obvious.
-
David Brownell authored
Getting some blood sugar made the real fix for that become obvious! This replaces the preceding TEMPORARY hack. Only unlink EDs that aren't already being unlinked! Duh.
-
David Brownell authored
Fix some OHCI TD list corruption issues: - Don't rewrite HC registers holding ED pointers until the HC had a good chance to finish using them. - Don't ever modify ed->hwTailP Adds text describing the different ED states. Adds TEMPORARY hack that may make a "rm_list becomes circular" bug continuable.
-
Linus Torvalds authored
This patch was trying to work around buggy SMM bios but causes problems by not deasserting NMI irq line (reported by Thomas Schlicht) for some integrated local apic. This patch was written originally as an optimization (with the side effect to fix those bugged bios) but was never applied to 2.4 tree. The rational was: fix your bios instead. Cset exclude: akpm@osdl.org|ChangeSet|20040219045328|16041
-
Andrew Morton authored
From: William Lee Irwin III <wli@holomorphy.com> Remember to invalidate the task->tty of threads, otherwise prod_pid_stat() later stumbles over the dangling pointers and crashes.
-
Linus Torvalds authored
The condition still seems to happen on ppc64.
-
Linus Torvalds authored
-
- 09 Mar, 2004 24 commits
-
-
Petr Vandrovec authored
sock_sendmsg() modifies iovec passed to it - it sets all length members of iovec array to zero on success transmission (and even on failed if it fails after iovec copy, but...) and advances pointers to point at the end of buffers used. This has an unfortunate effect that ncpfs's retry on failure does not work for IPX/UDP connections - kernel refused to do anything because length from iovec was 0 while length passed to sock_sendmsg() was correct. This simple fix gets rid of a problem by creating temporary iovec copy, which can sock_sendmsg destroy if it has such wish.
-
Gerd Knorr authored
This adds infrared remote support for a few more bt878-based TV cards.
-
Jens Axboe authored
From Mike Christie <michaelc@cs.wisc.edu> The first three bio and request flags are no longer identical. The bio barrier and rw flags are getting set in __make_request and get_request respectively, and failfast is getting left out. This sets the request's failfast flag in __make_request when the bio's flag is set.
-
Anton Blanchard authored
The ppc64 fix last week (enforcing permissions on the kernel when accessing userspace pages) uncovered a bug on POWER3/RS64. We werent zeroing the segment table entry before overwriting it and it was possible for the ks bit to be set on a kernel segment. The VSID mask was also changed to match reality (we only use 13 bits).
-
-
Richard Henderson authored
-
Benjamin Herrenschmidt authored
This patch reworks the early boot calls to OF to initialize displays. All present displays are now initialized in reverse order so the OF console stays on the first one. Initializing them all is necessary for dual head configurations as we need OF driver to properly setup the secondary TMDS of the video card, XFree isn't able to do that currently
-
Benjamin Herrenschmidt authored
The assembly code used to callback into Open Firmware client interface in 32 bits mode used to backup the stack pointer in the SPRG2 register. That upsets Apple's implementation of Open Firmware significantly and maybe others, causing them to crash in _some_ operations, apparently the trigger is to cause a segment or hash table fault, typically happens when letting that code initialize the second display. This patch fixes it, along with other cleanups of that asm code, it did unnecessary register restores and backing up the stack pointer is actually useless anyway.
-
Benjamin Herrenschmidt authored
Strangely, I though I fixed that a long time ago, but it was still broken in the current tree... Drivers like radeonfb fail to find the OF device matching a given PCI device on the G5 AGP bus because of some bus renumbering tricks. This patch fixes the problem by fixing the bus numbers in the OF node. This corrects radeonfb and other drivers looking for EDID / PLL datas in the OF node.
-
bk://kernel.bkbits.net/davem/net-2.6Linus Torvalds authored
into ppc970.osdl.org:/home/torvalds/v2.5/linux
-
Linus Torvalds authored
Especially with REGPARM these things actually matter.
-
Linus Torvalds authored
-
Benjamin Herrenschmidt authored
This fixes possible lockups in pmac_zilog when beeing flooded with incoming data (not that other serial drivers share the same race, I told Russel about it already). It also fixes some SCC initialization problems, add some PM callback, and fix the irda setup code.
-
Benjamin Herrenschmidt authored
This does cosmetic changes the pmac_zilog, changing the use of "up" as an identifier in favor "uap" to avoid collision with semaphores, use proper debug macros, and a few other cosmetic bits.
-
Andrew Morton authored
From Jan-Benedict Glaw <jbglaw@lug-owl.de>
-
Olaf Kirch authored
-
Krishna Kumar authored
-
Jean Tourrilhes authored
Original patch from Stephen Hemminger Move last bits of code out of irsyms and onto irmod.c where initialization happens. Move irda_debug parameter out of irsyms.c into irmod.c and make it a new style module parameter.
-
Stephen Hemminger authored
-
Stephen Hemminger authored
-
Stephen Hemminger authored
irda_device_setup can now be static, only called from alloc_irdadev.
-
Stephen Hemminger authored
-
Stephen Hemminger authored
-
Stephen Hemminger authored
Also, make type __u16 rather than unsigned short to match input parameter.
-