- 22 Jul, 2007 40 commits
-
-
Tim Hockin authored
Background: /dev/mcelog is typically polled manually. This is less than optimal for situations where accurate accounting of MCEs is important. Calling poll() on /dev/mcelog does not work. Description: This patch adds support for poll() to /dev/mcelog. This results in immediate wakeup of user apps whenever the poller finds MCEs. Because the exception handler can not take any locks, it can not call the wakeup itself. Instead, it uses a thread_info flag (TIF_MCE_NOTIFY) which is caught at the next return from interrupt or exit from idle, calling the mce_user_notify() routine. This patch also disables the "fake panic" path of the mce_panic(), because it results in printk()s in the exception handler and crashy systems. This patch also does some small cleanup for essentially unused variables, and moves the user notification into the body of the poller, so it is only called once per poll, rather than once per CPU. Result: Applications can now poll() on /dev/mcelog. When an error is logged (whether through the poller or through an exception) the applications are woken up promptly. This should not affect any previous behaviors. If no MCEs are being logged, there is no overhead. Alternatives: I considered simply supporting poll() through the poller and not using TIF_MCE_NOTIFY at all. However, the time between an uncorrectable error happening and the user application being notified is *the*most* critical window for us. Many uncorrectable errors can be logged to the network if given a chance. I also considered doing the MCE poll directly from the idle notifier, but decided that was overkill. Testing: I used an error-injecting DIMM to create lots of correctable DRAM errors and verified that my user app is woken up in sync with the polling interval. I also used the northbridge to inject uncorrectable ECC errors, and verified (printk() to the rescue) that the notify routine is called and the user app does wake up. I built with PREEMPT on and off, and verified that my machine survives MCEs. [wli@holomorphy.com: build fix] Signed-off-by:
Tim Hockin <thockin@google.com> Signed-off-by:
William Irwin <bill.irwin@oracle.com> Signed-off-by:
Andrew Morton <akpm@linux-foundation.org> Signed-off-by:
Andi Kleen <ak@suse.de> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
Tim Hockin authored
Background: /dev/mcelog is a clear-on-read interface. It is currently possible for multiple users to open and read() the device. Users are protected from each other during any one read, but not across reads. Description: This patch adds support for O_EXCL to /dev/mcelog. If a user opens the device with O_EXCL, no other user may open the device (EBUSY). Likewise, any user that tries to open the device with O_EXCL while another user has the device will fail (EBUSY). Result: Applications can get exclusive access to /dev/mcelog. Applications that do not care will be unchanged. Alternatives: A simpler choice would be to only allow one open() at all, regardless of O_EXCL. Testing: I wrote an application that opens /dev/mcelog with O_EXCL and observed that any other app that tried to open /dev/mcelog would fail until the exclusive app had closed the device. Caveats: None. Signed-off-by:
-
Aaron Durbin authored
Insert the unclaimed MMCONFIG resources into the resource tree without the IORESOURCE_BUSY flag during late initialization. This allows the MMCONFIG regions to be visible in the iomem resource tree without interfering with other system resources that were discovered during PCI initialization. [akpm@linux-foundation.org: nanofixes] Signed-off-by:
Aaron Durbin <adurbin@google.com> Cc: David Rientjes <rientjes@google.com> Signed-off-by:
-
David Rientjes authored
When we are in the emulated NUMA case, we need to make sure that all existing apicid_to_node mappings that point to real node ID's now point to the equivalent fake node ID's. If we simply iterate over all apicid_to_node[] members for each node, we risk remapping an entry if it shares a node ID with a real node. Since apicid's may not be consecutive, we're forced to create an automatic array of apicid_to_node mappings and then copy it over once we have finished remapping fake to real nodes. Signed-off-by:
David Rientjes <rientjes@google.com> Signed-off-by:
-
David Rientjes authored
For NUMA emulation, our SLIT should represent the true NUMA topology of the system but our proximity domain to node ID mapping needs to reflect the emulated state. When NUMA emulation has successfully setup fake nodes on the system, a new function, acpi_fake_nodes() is called. This function determines the proximity domain (_PXM) for each true node found on the system. It then finds which emulated nodes have been allocated on this true node as determined by its starting address. The node ID to PXM mapping is changed so that each fake node ID points to the PXM of the true node that it is located on. If the machine failed to register a SLIT, then we assume there is no special requirement for emulated node affinity so we use the default LOCAL_DISTANCE, which is newly exported to this code, as our measurement if the emulated nodes appear in the same PXM. Otherwise, we use REMOTE_DISTANCE. PXM_INVAL and NID_INVAL are also exported to the ACPI header file so that we can compare node_to_pxm() results in generic code (in this case, the SRAT code). Cc: Len Brown <lenb@kernel.org> Signed-off-by:
-
David Rientjes authored
The logic in e820_find_active_regions() for determining the true active regions for an e820 entry given a range of PFN's is needed for e820_hole_size() as well. e820_hole_size() is called from the NUMA emulation code to determine the reserved area within an address range on a per-node basis. Its logic should duplicate that of finding active regions in an e820 entry because these are the only true ranges we may register anyway. [akpm@linux-foundation.org: cleanup] Cc: Mel Gorman <mel@csn.ul.ie> Signed-off-by:
-
Christoph Lameter authored
This adds caching of pgds and puds, pmds, pte. That way we can avoid costly zeroing and initialization of special mappings in the pgd. A second quicklist is useful to separate out PGD handling. We can carry the initialized pgds over to the next process needing them. Also clean up the pgd_list handling to use regular list macros. There is no need anymore to avoid the lru field. Move the add/removal of the pgds to the pgdlist into the constructor / destructor. That way the implementation is congruent with i386. Signed-off-by:
Christoph Lameter <clameter@sgi.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: "Luck, Tony" <tony.luck@intel.com> Acked-by:
William Lee Irwin III <wli@holomorphy.com> Signed-off-by:
-
Adrian Bunk authored
timer_irq_works() needlessly became global. Signed-off-by:
Adrian Bunk <bunk@stusta.de> Signed-off-by:
-
Adrian Bunk authored
Every file should include the headers containing the prototypes for its global functions. Signed-off-by:
-
Adrian Bunk authored
Signed-off-by:
-
Jan Beulich authored
Constrain __supported_pte_mask and NX handling to just the PAE kernel. Signed-off-by:
Jan Beulich <jbeulich@novell.com> Signed-off-by:
-
Jan Beulich authored
Hence remove its handling in the opposite case. Signed-off-by:
-
Jan Beulich authored
.. and adjust documentation to properly reflect options that are x86-64 specific. Signed-off-by:
-
Jan Beulich authored
Signed-off-by:
-
Jan Beulich authored
Signed-off-by:
-
Jan Beulich authored
Consolidate the three 32-bit system call entry points so that they all treat registers in similar ways. Signed-off-by:
-
Ravikiran G Thirumalai authored
Too many remote cpu references due to /proc/stat. On x86_64, with newer kernel versions, kstat_irqs is a bit of a problem. On every call to kstat_irqs, the process brings in per-cpu data from all online cpus. Doing this for NR_IRQS, which is now 256 + 32 * NR_CPUS results in (256+32*63) * 63 remote cpu references on a 64 cpu config. /proc/stat is parsed by common commands like top, who etc, causing lots of cacheline transfers This statistic seems useless. Other 'big iron' arches disable this. AK: changed to remove for all SMP setups AK: add comment Signed-off-by:
Ravikiran Thirumalai <kiran@scalex86.org> Signed-off-by:
-
Thomas Gleixner authored
Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
Chris Wright <chrisw@sous-sol.org> Signed-off-by:
Ingo Molnar <mingo@elte.hu> Signed-off-by:
-
Thomas Gleixner authored
Fix coding style, white space wreckage and remove unused code. Signed-off-by:
-
Thomas Gleixner authored
hpet.h in asm-i386 and asm-x86_64 contain tons of duplicated stuff. Consolidate into one shared header file. AK: Fix i386 compilation with !X86_IO_APIC Signed-off-by:
-
Thomas Gleixner authored
The hpet_rtc_interrupt handler still uses pt_regs. Fix it. Signed-off-by:
-
Thomas Gleixner authored
Signed-off-by:
-
Thomas Gleixner authored
Remove unused code and variables and do some codingstyle / whitespace cleanups while at it. Cc: john stultz <johnstul@us.ibm.com> Signed-off-by:
-
Thomas Gleixner authored
xtime can be initialized including the cmos update from the generic timekeeping code. Remove the arch specific implementation. Signed-off-by:
-
Thomas Gleixner authored
Use the generic cmos update function in kernel/time/ntp.c Cc: john stultz <johnstul@us.ibm.com> Signed-off-by:
-
Chris Wright authored
When making changes to x86_64 timers, I noticed that touching hpet.h triggered an unreasonably large rebuild. Untangling it from timex.h quiets the extra rebuild quite a bit. Cc: john stultz <johnstul@us.ibm.com> Signed-off-by:
-
Chris Wright authored
Remove pit_interrupt_hook as it adds just an extra layer. Signed-off-by:
-
Thomas Gleixner authored
The current SMI detection logic in read_hpet_tsc() makes sure, that when a SMI happens between the read of the HPET counter and the read of the TSC, this wrong value is used for TSC calibration. This is not the intention of the function. The comparison must ensure, that we do _NOT_ use such a value. Fix the check to use calibration values where delta of the two TSC reads is smaller than a reasonable threshold. Signed-off-by:
-
Björn Steinbrink authored
The Intel PerfMon NMI watchdog reserves the first performance counter, but uses the second one. Make it correctly reserve the second one. Signed-off-by:
Björn Steinbrink <B.Steinbrink@gmx.de> Signed-off-by:
-
Andi Kleen authored
It is not fully softirq safe anyways. Can't do a WARN_ON unfortunately because it could trigger in the panic case. Signed-off-by:
-
Yinghai Lu authored
Signed-off-by:
Yinghai Lu <yinghai.lu@sun.com> Signed-off-by:
-
Andi Kleen authored
With that an L3 cache is correctly reported in the cache information in /sys With fixes from Andreas Herrmann and Dean Gaudet and Joachim Deguara Signed-off-by:
-
Andi Kleen authored
This implements new vDSO for x86-64. The concept is similar to the existing vDSOs on i386 and PPC. x86-64 has had static vsyscalls before, but these are not flexible enough anymore. A vDSO is a ELF shared library supplied by the kernel that is mapped into user address space. The vDSO mapping is randomized for each process for security reasons. Doing this was needed for clock_gettime, because clock_gettime always needs a syscall fallback and having one at a fixed address would have made buffer overflow exploits too easy to write. The vdso can be disabled with vdso=0 It currently includes a new gettimeofday implemention and optimized clock_gettime(). The gettimeofday implementation is slightly faster than the one in the old vsyscall. clock_gettime is significantly faster than the syscall for CLOCK_MONOTONIC and CLOCK_REALTIME. The new calls are generally faster than the old vsyscall. Advantages over the old x86-64 vsyscalls: - Extensible - Randomized - Cleaner - Easier to virtualize (the old static address range previously causes overhead e.g. for Xen because it has to create special page tables for it) Weak points: - glibc support still to be written The VM interface is partly based on Ingo Molnar's i386 version. Includes compile fix from Joachim Deguara Signed-off-by:
-
Andi Kleen authored
gcc 4.3 supports a new __attribute__((__cold__)) to mark functions cold. Any path directly leading to a call of this function will be unlikely. And gcc will try to generate smaller code for the function itself. Please use with care. The code generation advantage isn't large and in most cases it is not worth uglifying code with this. This patch marks some common error functions like panic(), printk() as cold. This will longer term make many unlikely()s unnecessary, although we can keep them for now for older compilers. BUG is not marked cold because there is currently no way to tell gcc to mark a inline function told. Also all __init and __exit functions are marked cold. With a non -Os build this will tell the compiler to generate slightly smaller code for them. I think it currently only uses less alignments for labels, but that might change in the future. One disadvantage over *likely() is that they cannot be easily instrumented to verify them. Another drawback is that only the latest gcc 4.3 snapshots support this. Unfortunately we cannot detect this using the preprocessor. This means older snapshots will fail now. I don't think that's a problem because they are unreleased compilers that nobody should be using. gcc also has a __hot__ attribute, but I don't see any sense in using this in the kernel right now. But someday I hope gcc will be able to use more aggressive optimizing for hot functions even in -Os, if that happens it should be added. Includes compile fix from Thomas Gleixner. Cc: Jan Hubicka <jh@suse.cz> Signed-off-by:
-
Andi Kleen authored
The compiler generally generates reasonable inline code for the simple cases and for the rest it's better for code size for them to be out of line. Also there they can be potentially optimized more in the future. In fact they probably should be in a .S file because they're all pure assembly, but that's for another day. Also some code style cleanup on them while I was on it (this seems to be the last untouched really early Linux code) This saves ~12k text for a defconfig kernel with gcc 4.1. Signed-off-by:
-
Andi Kleen authored
Jan asked to always use the builtin memcpy on gcc 4.3 mainline because it should generate better code than the old macro. Let's try it. Cc: Jan Hubicka <jh@suse.cz> Signed-off-by:
-
Andi Kleen authored
Signed-off-by:
-
David Rientjes authored
In acpi_scan_nodes(), we immediately return -1 if acpi_numa <= 0, meaning we haven't detected any underlying ACPI topology or we have explicitly disabled its use from the command-line with numa=noacpi. acpi_table_print_srat_entry() and acpi_table_parse_srat() are only referenced within drivers/acpi/numa.c, so we can mark them as static and remove their prototypes from the header file. Likewise, pxm_to_node_map[] and node_to_pxm_map[] are only used within drivers/acpi/numa.c, so we mark them as static and remove their externs from the header file. The automatic 'result' variable is unused in acpi_numa_init(), so it's removed. Signed-off-by:
-
David Rientjes authored
Use LOCAL_DISTANCE and REMOTE_DISTANCE in x86_64 ACPI code Signed-off-by:
-
Andi Kleen authored
Otherwise smp_affinity would only update after the next interrupt on x86 systems. Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Ingo Molnar <mingo@elte.hu> Signed-off-by:
-
