- 30 Jan, 2008 40 commits
-
-
Yinghai Lu authored
Andi's patch " x86: move X86_FEATURE_CONSTANT_TSC into early cpu feature detection Need this in the next patch in time_init and that happens early. This includes a minor fix on i386 where early_intel_workarounds() [which is now called early_init_intel] really executes early as the comments say. " calling early_init_amd in early_identify_cpu and identify_cpu two times. this patch remove the one in identify_cpu Signed-off-by:Yinghai Lu <yinghai.lu@sun.com> Signed-off-by:
Ingo Molnar <mingo@elte.hu> Signed-off-by:
Thomas Gleixner <tglx@linutronix.de>
-
Jesse Barnes authored
On some machines, buggy BIOSes don't properly setup WB MTRRs to cover all available RAM, meaning the last few megs (or even gigs) of memory will be marked uncached. Since Linux tends to allocate from high memory addresses first, this causes the machine to be unusably slow as soon as the kernel starts really using memory (i.e. right around init time). This patch works around the problem by scanning the MTRRs at boot and figuring out whether the current end_pfn value (setup by early e820 code) goes beyond the highest WB MTRR range, and if so, trimming it to match. A fairly obnoxious KERN_WARNING is printed too, letting the user know that not all of their memory is available due to a likely BIOS bug. Something similar could be done on i386 if needed, but the boot ordering would be slightly different, since the MTRR code on i386 depends on the boot_cpu_data structure being setup. This patch fixes a bug in the last patch that caused the code to run on non-Intel machines (AMD machines apparently don't need it and it's untested on other non-Intel machines, so best keep it off). Further enhancements and fixes from: Yinghai Lu <Yinghai.Lu@Sun.COM> Andi Kleen <ak@suse.de> Signed-off-by:
Jesse Barnes <jesse.barnes@intel.com> Tested-by:
Justin Piszcz <jpiszcz@lucidpixels.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: "Eric W. Biederman" <ebiederm@xmission.com> Cc: Yinghai Lu <yhlu.kernel@gmail.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
-
Andi Kleen authored
They now look like: hal-resmgr[13791]: segfault at 3c rip 2b9c8caec182 rsp 7fff1e825d30 error 4 in libacl.so.1.1.0[2b9c8caea000+6000] This makes it easier to pinpoint bugs to specific libraries. And printing the offset into a mapping also always allows to find the correct fault point in a library even with randomized mappings. Previously there was no way to actually find the correct code address inside the randomized mapping. Relies on earlier patch to shorten the printk formats. They are often now longer than 80 characters, but I think that's worth it. [includes fix from Eric Dumazet to check d_path error value] Signed-off-by:
Andi Kleen <ak@suse.de> Signed-off-by:
-
Andi Kleen authored
When the kernel panics early for some unrelated reason there would be eventually an early exception inside panic because clear_local_APIC tried to disable the not yet mapped APIC. Check for that explicitely. Signed-off-by:
-
Andi Kleen authored
On VMs implemented using JITs that cache translated code changing the lock prefixes is a quite costly operation that forces the JIT to throw away and retranslate a lot of code. Previously a SMP kernel would rewrite the locks once for each CPU which is quite unnecessary. This patch changes the code to never switch at boot in the normal case (SMP kernel booting with >1 CPU) or only once for SMP kernel on UP. This makes a significant difference in boot up performance on AMD SimNow! Also I expect it to be a little faster on native systems too because a smp switch does a lot of text_poke()s which each synchronize the pipeline. v1->v2: Rename max_cpus v1->v2: Fix off by one in UP check (Thomas Gleixner) Signed-off-by:
-
Andi Kleen authored
On x86-64 there are several memory allocations before bootmem. To avoid them stomping on each other they used to be all hard coded in bad_area(). Replace this with an array that is filled as needed. This cleans up the code considerably and allows to expand its use. Cc: peterz@infradead.org Signed-off-by:
-
Harvey Harrison authored
Adding the address of the faulting library missed removing a line ending from X86_32. Also update the shorter printk format for X86_32 in fault_64.c to make it easier to se the remaining differences. Signed-off-by:
Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
Andi Kleen authored
Previously there was a AMD specific quirk to handle the case of AMD Fam10h MWAIT not supporting any C states. But it turns out that CPUID already has ways to detectly detect that without using special quirks. The new code simply checks if MWAIT supports at least C1 and doesn't use it if it doesn't. No more vendor specific code. Note this is does not simply clear MWAIT because MWAIT can be still useful even without C states. Credit goes to Ben Serebrin for pointing out the (nearly) obvious. Cc: "Andreas Herrmann" <andreas.herrmann3@amd.com> Signed-off-by:
-
Andi Kleen authored
Previously it was only run for Intel CPUs, but AMD Fam10h implements MWAIT too. This matches 64bit behaviour. Signed-off-by:
-
Harvey Harrison authored
Choose a less generic name for such a special case. Add a comment explaining the odd use in X86_32. Change the one user of stack_pointer. Signed-off-by:
-
Harvey Harrison authored
Leave definition of pt_regs in its own section, move all kernel code to section afterwards, unify prototype definitions, has some conditional prototypes to make it clear what was only defined in 32 and 64 bit. Signed-off-by:
-
Harvey Harrison authored
Unify the definiton of: v8086_mode user_mode user_mode_vm stack_pointer instruction_pointer frame_pointer in ptrace.h to make it clear where the differences are between 32 and 64 bit. Changes macros to static inlines as well. Signed-off-by:
-
Hiroshi Shimamoto authored
kdump needs ELF_CORE_COPY_REGS in crash_save_cpu(). This lack of the macro causes the following BUG. SysRq : Trigger a crashdump ------------[ cut here ]------------ kernel BUG at include/linux/elfcore.h:105! invalid opcode: 0000 [1] PREEMPT SMP Signed-off-by:
Hiroshi Shimamoto <h-shimamoto@ct.jp.nec.com> Signed-off-by:
-
Dmitri Vorobiev authored
This is against current x86.git. The size of the system call table for 32-bit x86 kernels is obtained by compile-time calculation of the sys_call_table array, not from the value, which the NR_syscalls macro expands to. This trivial patch removes the fossil macro. Manually tested by grepping the x86 files for the "NR_syscalls" string. No relevant use cases found. Build-tested using allyesconfig, allnoconfig and a couple of randconfig instances. All builds successfully finished. Runtime test performed using a stripped-down Debian-ish config. The system booted successfully. Signed-off-by:
Dmitri Vorobiev <dmitri.vorobiev@gmail.com> Signed-off-by:
-
Kyle McMartin authored
PSE, PGE, XMM, XMM2, and FXSR are defined as required features, and will be optimized to a constant at compile time. Remove their redundant definitions. Signed-off-by:
Kyle McMartin <kyle@mcmartin.ca> Signed-off-by:
-
travis@sgi.com authored
This patch removes the EXPORT_SYMBOL for: x86_cpu_to_node_map_init x86_cpu_to_node_map_early_ptr ... thus fixing the section mismatch problem. Also, the mem -> node hash lookup is fixed. Signed-off-by:
Mike Travis <travis@sgi.com> Signed-off-by:
-
Jeremy Fitzhardinge authored
Remove duplicate set_pud()s. Signed-off-by:
Jeremy Fitzhardinge <jeremy@xensource.com> Signed-off-by:
-
Jeremy Fitzhardinge authored
Remove duplicate set_pmd()s. Signed-off-by:
-
Jeremy Fitzhardinge authored
Remove duplicate set_pte* operations. PAE still needs to have special variants of some of these because it can't atomically update a 64-bit pte, so there's still some duplication. Signed-off-by:
-
Jeremy Fitzhardinge authored
Remove duplicate __pmd/pmd_val functions. Signed-off-by:
-
Jeremy Fitzhardinge authored
Remove duplicate __pgd/pgd_val functions. Signed-off-by:
-
Jeremy Fitzhardinge authored
Remove duplicate __pte/pte_val functions. Signed-off-by:
-
Jeremy Fitzhardinge authored
Rearrange the various pagetable mmu_ops to remove duplication. Signed-off-by:
-
Jeremy Fitzhardinge authored
Signed-off-by:
-
Andrew Morton authored
WARNING: vmlinux.o(__ksymtab+0x670): Section mismatch: reference to .init.data:x86_cpu_to_node_map_init (between '__ksymtab_x86_cpu_to_node_map_init' and '__ksymtab_node_data') Cc: Matthew Dobson <colpatch@us.ibm.com> Cc: Mike Travis <travis@sgi.com> Cc: Christoph Lameter <clameter@sgi.com> Signed-off-by:
-
Mike Travis authored
Signed-off-by:
-
Jan Beulich authored
Based on patch from Jan Beulich <jbeulich@novell.com>. Don't rely on kmalloc(PAGE_SIZE) returning PAGE_SIZE aligned memory (Xen requires GDT *and* LDT to be page-aligned). Using the page allocator interface also removes the (albeit small) slab allocator overhead. The same change being done for 64-bits for consistency. Further, the Xen hypercall interface expects the LDT address to be virtual, not machine. [ Adjusted to unified ldt.c - Jeremy ] Signed-off-by:
Jan Beulich <jbeulich@novell.com> Acked-by:
-
Jan Beulich authored
Remove the dead .text.lock. Move _etext and __{start,stop}___ex_table into their sections. Signed-off-by: -
Alan Cox authored
I can find no reason for the _p on the serverworks IRQ routing logic, and a review of the documentation contains no indication that any such delay is needed so lets try this Signed-off-by:
-
Alan Cox authored
Rather than remove and/or mangle inb_p/outb_p we want to remove the use of them from inappropriate places. For the PIC/PIT this may eventually depend on 32/64bitism or similar so start by adding inb/outb_pit and inb/outb_pic so that we can make them use any scheme we settle on without disturbing the existing, correct (for ISA), port 0x80 usage. (eg we can make inb_pit use udelay without messing up inb_p). Floppy already does this for the fdc. That really only leaves the CMOS as a core logic item to tackle, and bits of parallel port handling in the chipset layers. Signed-off-by:
Alan Cox <alan@redhat.com> Signed-off-by:
-
Abhishek Sagar authored
Highlight peculiar cases in singles-step kprobe handling. In reenter_kprobe(), a breakpoint in KPROBE_HIT_SS case can only occur when single-stepping a breakpoint on which a probe was installed. Since such probes are single-stepped inline, identifying these cases is unambiguous. All other cases leading up to KPROBE_HIT_SS are possible bugs. Identify and WARN_ON such cases. Signed-off-by:
Abhishek Sagar <sagar.abhishek@gmail.com> Signed-off-by:
-
Benjamin LaHaise authored
an otherwise idle system takes about 3 ticks per network interface in unregister_netdev() due to multiple calls to synchronize_rcu(), which adds up to quite a few seconds for tearing down thousands of interfaces. By flushing pending rcu callbacks in the idle loop, the system makes progress hundreds of times faster. If this is indeed a sane thing to, it probably needs to be done for other architectures than x86. And yes, the network stack shouldn't call synchronize_rcu() quite so much, but fixing that is a little more involved. Signed-off-by:
-
John Reiser authored
The ENDPROCs() were not used everywhere. Some code used just END() instead, while other code used nothing. um/sys-i386/checksum.S didn't #include <linux/linkage.h> . I also got confused because gcc puts the .type near the ENTRY, while ENDPROC puts it on the opposite end. Signed off by: John Reiser <jreiser@BitWagon.com> Signed-off-by:
-
Harvey Harrison authored
Add caller of is_errata93() to X86_32, ifdef'd to do nothing. Signed-off-by:
-
Harvey Harrison authored
Comments, indentation, printk format. Uses task_pid_nr() on X86_64 now, but this is always defined to task->pid. Signed-off-by:
-
Harvey Harrison authored
Copy the prefetch of map_sem from X86_64 and move the check notify_page_fault (soon to be kprobe_handle_fault) out of the unlikely if() statement. This makes the X86_32|64 pagefault handlers closer to each other. Signed-off-by:
-
Harvey Harrison authored
is_prefetch was the last user of get_segment_eip and only on X86_32. This function returned the faulting instruction's address and set the upper segment limit. Instead, use the convert_ip_to_linear helper and rely on probe_kernel_address to do the segment checks which was already done everywhere the segment limit was being checked on X86_32. Remove get_segment_eip as well. Signed-off-by:
-
Harvey Harrison authored
Rename convert_rip_to_linear to convert_ip_to_linear for shared X86_32|64 use. Signed-off-by:
-
Harvey Harrison authored
Where x86_32 passed zero in the high 32 bits, use wrmsrl which will zero extend for us. This allows ifdefs for 32/64 bit to be eliminated. Eliminate ifdef in step.c. Similar cleanup was done when unifying kprobes_32|64.c and wrmsr() was chosen there over wrmsrl(). This patch changes these to wrmsrl. Signed-off-by:
-
