- 26 Jun, 2006 40 commits
-
-
Keith Owens authored
On some i386/x86_64 systems, sending an NMI IPI as a broadcast will reset the system. This seems to be a BIOS bug which affects machines where one or more cpus are not under OS control. It occurs on HT systems with a version of the OS that is not compiled without HT support. It also occurs when a system is booted with max_cpus=n where 2 <= n < cpus known to the BIOS. The fix is to always send NMI IPI as a mask instead of as a broadcast. Signed-off-by:
Keith Owens <kaos@sgi.com> Signed-off-by:
Andi Kleen <ak@suse.de> Signed-off-by:
Linus Torvalds <torvalds@osdl.org>
-
Siddha, Suresh B authored
Appended patch fixes the "APIC error on CPUX: 00(40)" observed during bootup. From SDM Vol-3A "Valid Interrupt Vectors" section: "When an illegal vector value (0-15) is written to an LVT entry and the delivery mode is Fixed, the APIC may signal an illegal vector error, with out regard to whether the mask bit is set or whether an interrupt is actually seen on input." Signed-off-by:
Suresh Siddha <suresh.b.siddha@intel.com> Signed-off-by:
-
Chuck Ebbert authored
Allow stack growth so the 'enter' instruction works. Also fixes problem in compat_sys_kexec_load() which could allocate more than 128 bytes using compat_alloc_user_space(). Signed-off-by:
Chuck Ebbert <76306.1226@compuserve.com> Signed-off-by:
-
Andi Kleen authored
- Use tail call from clear_user to __clear_user to save some code size - Use standard memcpy for forward memmove Signed-off-by:
-
Andi Kleen authored
Only exports for assembler files are left in x8664_ksyms.c Originally inspired by a patch from Al Viro Signed-off-by:
-
Keith Owens authored
x86_64 and i386 behave inconsistently when sending an IPI on vector 2 (NMI_VECTOR). Make both behave the same, so IPI 2 is sent as NMI. The crash code was abusing send_IPI_allbutself() by passing a code instead of a vector, it only worked because crash knew about the internal code of send_IPI_allbutself(). Change crash to use NMI_VECTOR instead, and remove the comment about how crash was abusing the function. This patch is a pre-requisite for fixing the problem where sending an IPI as NMI would reboot some Dell Xeon systems. I cannot fix that problem while crash continus to abuse send_IPI_allbutself(). It also removes the inconsistency between i386 and x86_64 for NMI_VECTOR. That will simplify all the RAS code that needs to bring all the cpus to a clean stop, even when one or more cpus are spinning disabled. Signed-off-by:
-
Piotr Kaczuba authored
It turned out that the following change is needed when the speaker is compiled as a module. Signed-off-by:
-
Adrian Bunk authored
This patch removes the no longer used sys32_ni_syscall() Signed-off-by:
Adrian Bunk <bunk@stusta.de> Signed-off-by:
-
Sergey Vlasov authored
Currently CONFIG_REORDER uses -ffunction-sections for all code; however, creating a separate section for each function is not useful for modules and just adds bloat. Moving this option from CFLAGS to CFLAGS_KERNEL shrinks module object files (e.g., the module tree for a kernel built with most drivers as modules shrinked from 54M to 46M), and decreases the number of sysfs files in /sys/module/*/sections/ directories. Signed-off-by:
Sergey Vlasov <vsu@altlinux.ru> Signed-off-by:
-
Andreas Mohr authored
This one is adding a cpu_relax() that already existed in the i386 version. Signed-off-by:
Andreas Mohr <andi@lisas.de> Signed-off-by:
-
Jan Beulich authored
Defaulting to a value not evenly divisible by four makes little sense, as four values are displayed per line (and hence the rest of the line would otherwise be wasted). Signed-off-by:
Jan Beulich <jbeulich@novell.com> Signed-off-by:
-
Jan Beulich authored
With (significantly) more than 10 CPUs online, the column headings drifted off the positions of the column contents with growing CPU numbers. Signed-off-by:
-
Andi Kleen authored
When a process changes CPUs while doing the non atomic cpu_local_* operations it might operate on the local_t of a different CPUs. Fix that by disabling preemption. Pointed out by Christopher Lameter Signed-off-by:
-
Andi Kleen authored
The APIC ID returned by hard_smp_processor_id can be beyond NR_CPUS and then overflow the x86_cpu_to_apic[] array. Add a check for overflow. If it happens then the slow loop below will catch. Bug pointed out by Doug Thompson Signed-off-by:
-
Rohit Seth authored
Getting phys_proc_id and cpu_core_id information to be printed at boot time for AMD processors. Also matching the Node related boot time information that gets printed for Intel and AMD processors for NUMA configurations. Signed-off-by:
Rohit Seth <rohitseth@google.com> Signed-off-by:
-
Andi Kleen authored
During some profiling I noticed that default_idle causes a lot of memory traffic. I think that is caused by the atomic operations to clear/set the polling flag in thread_info. There is actually no reason to make this atomic - only the idle thread does it to itself, other CPUs only read it. So I moved it into ti->status. Converted i386/x86-64/ia64 for now because that was the easiest way to fix ACPI which also manipulates these flags in its idle function. Cc: Nick Piggin <npiggin@novell.com> Cc: Tony Luck <tony.luck@intel.com> Cc: Len Brown <len.brown@intel.com> Signed-off-by:
-
Andi Kleen authored
No red zone possible/needed on the alternative stack. It caused confusion. Signed-off-by:
-
Andrew Morton authored
- Use DMA_32BIT_MASK - Use %z for size_t - 80-cols Cc: Andi Kleen <ak@suse.de> Signed-off-by:
Andrew Morton <akpm@osdl.org> Signed-off-by:
-
Jan Beulich authored
- fix an off-by-one error in phys_pmd_init() - prevent phys_pmd_init() from removing mappings established earlier - fix the direct mapping early printk to in fact show the end of the range - remove an apparently orphan comment Signed-off-by:
-
Andi Kleen authored
Cc: Jacob Shin <jacob.shin@amd.com> Signed-off-by:
-
Jacob Shin authored
Clean up mce_amd.c for readability and remove code no longer needed. Signed-off-by:
Jacob Shin <jacob.shin@amd.com> Signed-off-by:
-
Jacob Shin authored
Add support for mce threshold registers found in future AMD family 0x10 processors. Backwards compatible with family 0xF hardware. AK: fixed build on !SMP Signed-off-by:
-
Jacob Shin authored
Get rid of /sys/devices/system/threshold directory and move mce_amd thresholding files into the machine sysfs directory -- /sys/devices/system/machinecheck. AK: Fixed warning Signed-off-by:
-
Jacob Shin authored
Add support for extended APIC LVT found in future AMD processors. Signed-off-by:
-
Andi Kleen authored
Architecture specific configs like this have no business at all in init/Kconfig. This prevents it from being set on x86-64 Pointed out by H.Peter Anvin Signed-off-by:
-
Andi Kleen authored
It's like this on SUSE systems. Cc: hpa@zytor.com Signed-off-by:
-
Vojtech Pavlik authored
Update my copyright dates in arch/x86-64/kernel/time.c Signed-off-by:
Vojtech Pavlik <vojtech@suse.cz> Signed-off-by:
-
Vojtech Pavlik authored
After writing the CFG register, the first value written to the T0_CMP register is the value at which next interrupt should be triggered, every value after that sets the period of the interrupt. For that reason, the code needs to write the value twice - to set both the phase and period. [AK: I had already figured it out by myself, but it's still useful to have a comment for this.] Signed-off-by:
-
Vojtech Pavlik authored
This patch makes use of the newly added conversion constants in time.h to x86-64 time.c. The code gets significantly easier to understand. Signed-off-by:
-
Vojtech Pavlik authored
Remove #ifdefed code to manually enable HPET on AMD8111, where the BIOS doesn't have ACPI HPET tables and doesn't enable it for us. Signed-off-by:
-
Vojtech Pavlik authored
This patch adds the X86_FEATURE_RDTSCP #define, so that kernel code can check for the feature easily and also fixes the location of the "rdtscp" string in the cpuinfo tables. Signed-off-by:
-
Vojtech Pavlik authored
Rename oem_force_hpet_timer to apic_is_clustered_box, to give the function a better fitting name - it really isn't at all about HPET. Signed-off-by:
-
Vojtech Pavlik authored
In timekeeping code, one often does need to use conversion constants. Naming these leads to code that's easier to understand, showing the reader between which units the conversion is made. Signed-off-by:
-
Rohit Seth authored
Most of the fields of cpuinfo are defined in cpuinfo_x86 structure. This patch moves the phys_proc_id and cpu_core_id for each processor to cpuinfo_x86 structure as well. Signed-off-by:
-
Jon Mason authored
This patch hooks Calgary into the build, the x86-64 IOMMU initialization paths, and introduces the Calgary specific bits. The implementation draws inspiration from both PPC (which has support for the same chip but requires firmware support which we don't have on x86-64) and gart. Calgary is different from gart in that it support a translation table per PHB, as opposed to the single gart aperture. Changes from previous version: * Addition of boot-time disablement for bus-level translation/isolation (e.g, enable userspace DMA for things like X) * Usage of newer IOMMU abstraction functions Signed-off-by:
Muli Ben-Yehuda <muli@il.ibm.com> Signed-off-by:
Jon Mason <jdmason@us.ibm.com> Signed-off-by:
-
Jon Mason authored
This patch creates a new interface for IOMMUs by adding a centralized location for IOMMU allocation (for translation tables/apertures) and IOMMU initialization. In creating these, code was moved around for abstraction, uniformity, and consiceness. Take note of the move of the iommu_setup bootarg parsing code to __setup. This is enabled by moving back the location of the aperture allocation/detection to mem init (which while ugly, was already the location of the swiotlb_init). While a slight departure from the previous patch, I belive this provides the true intention of the previous versions of the patch which changed this code. It also makes the addition of the upcoming calgary code much cleaner than previous patches. [AK: Removed one broken change. iommu_setup still has to be called early] Signed-off-by:
-
Jon Mason authored
Based on Andi Kleen's comments on the original Calgary patch, move valid_dma_direction into the calling functions. Signed-off-by:
-
Jon Mason authored
swiotlb relies on the gart specific iommu_aperture variable to know if we discovered a hardware IOMMU before swiotlb initialization. Introduce iommu_detected to do the same thing, but in a HW IOMMU neutral manner, in preparation for adding the Calgary HW IOMMU. Signed-Off-By:
-
Rohit Seth authored
Minor cleanup patch: Replacing the asm statement with cpuid_count macro(which already provides the same functionality). Signed-off-by:
-
Jan Beulich authored
Use abstractions whenever possible. Signed-off-by:
-
