)]}'
{
  "log": [
    {
      "commit": "bd78432c8f209a1028f4e5bada8b1da1d8e4da09",
      "tree": "d807bc0f28f3157927f93c37c3e5002b00633b2e",
      "parents": [
        "e02e68d31e2d436197386997727b216fee9c4623"
      ],
      "author": {
        "name": "Tim Hockin",
        "email": "thockin@google.com",
        "time": "Sat Jul 21 17:10:37 2007 +0200"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@woody.linux-foundation.org",
        "time": "Sat Jul 21 18:37:10 2007 -0700"
      },
      "message": "x86_64: mcelog tolerant level cleanup\n\nBackground:\n The MCE handler has several paths that it can take, depending on various\n conditions of the MCE status and the value of the \u0027tolerant\u0027 knob.  The\n exact semantics are not well defined and the code is a bit twisty.\n\nDescription:\n This patch makes the MCE handler\u0027s behavior more clear by documenting the\n behavior for various \u0027tolerant\u0027 levels.  It also fixes or enhances\n several small things in the handler.  Specifically:\n     * If RIPV is set it is not safe to restart, so set the \u0027no way out\u0027\n       flag rather than the \u0027kill it\u0027 flag.\n     * Don\u0027t panic() on correctable MCEs.\n     * If the _OVER bit is set *and* the _UC bit is set (meaning possibly\n       dropped uncorrected errors), set the \u0027no way out\u0027 flag.\n     * Use EIPV for testing whether an app can be killed (SIGBUS) rather\n       than RIPV.  According to docs, EIPV indicates that the error is\n       related to the IP, while RIPV simply means the IP is valid to\n       restart from.\n     * Don\u0027t clear the MCi_STATUS registers until after the panic() path.\n       This leaves the status bits set after the panic() so clever BIOSes\n       can find them (and dumb BIOSes can do nothing).\n\n This patch also calls nonseekable_open() in mce_open (as suggested by akpm).\n\nResult:\n Tolerant levels behave almost identically to how they always have, but\n not it\u0027s well defined.  There\u0027s a slightly higher chance of panic()ing\n when multiple errors happen (a good thing, IMHO).  If you take an MBE and\n panic(), the error status bits are not cleared.\n\nAlternatives:\n None.\n\nTesting:\n I used software to inject correctable and uncorrectable errors.  With\n tolerant \u003d 3, the system usually survives.  With tolerant \u003d 2, the system\n usually panic()s (PCC) but not always.  With tolerant \u003d 1, the system\n always panic()s.  When the system panic()s, the BIOS is able to detect\n that the cause of death was an MC4.  I was not able to reproduce the\n case of a non-PCC error in userspace, with EIPV, with (tolerant \u003c 3).\n That will be rare at best.\n\nSigned-off-by: Tim Hockin \u003cthockin@google.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n"
    },
    {
      "commit": "d567b6a955c2c1a6b0e8fc13ecddc7d4ac4900a2",
      "tree": "7920c076cc3dae5684c47e8c7d7b002247d0fe47",
      "parents": [
        "74a1ddc597d83d5ca6541b15a15f80a3722e3a1e"
      ],
      "author": {
        "name": "Jan Beulich",
        "email": "jbeulich@novell.com",
        "time": "Sat Jul 21 17:10:23 2007 +0200"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@woody.linux-foundation.org",
        "time": "Sat Jul 21 18:37:09 2007 -0700"
      },
      "message": "x86_64: remove unused variable maxcpus\n\n.. and adjust documentation to properly reflect options that are\nx86-64 specific.\n\nSigned-off-by: Jan Beulich \u003cjbeulich@novell.com\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n"
    },
    {
      "commit": "8a336b0a4b6dfacc8cc5fd617ba1e1904077de2d",
      "tree": "9e2a5f4aeb080fe68a1cf26860ebc7f69e9fccb2",
      "parents": [
        "f82af20e1a028e16b9bb11da081fa1148d40fa6a"
      ],
      "author": {
        "name": "Tim Hockin",
        "email": "thockin@google.com",
        "time": "Wed May 02 19:27:19 2007 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Wed May 02 19:27:19 2007 +0200"
      },
      "message": "[PATCH] x86-64: Dynamically adjust machine check interval\n\nBackground:\n We\u0027ve found that MCEs (specifically DRAM SBEs) tend to come in bunches,\n especially when we are trying really hard to stress the system out.  The\n current MCE poller uses a static interval which does not care whether it\n has or has not found MCEs recently.\n\nDescription:\n This patch makes the MCE poller adjust the polling interval dynamically.\n If we find an MCE, poll 2x faster (down to 10 ms).  When we stop finding\n MCEs, poll 2x slower (up to check_interval seconds).  The check_interval\n tunable becomes the max polling interval.  The \"Machine check events\n logged\" printk() is rate limited to the check_interval, which should be\n identical behavior to the old functionality.\n\nResult:\n If you start to take a lot of correctable errors (not exceptions), you\n log them faster and more accurately (less chance of overflowing the MCA\n registers).  If you don\u0027t take a lot of errors, you will see no change.\n\nAlternatives:\n I considered simply reducing the polling interval to 10 ms immediately\n and keeping it there as long as we continue to find errors.  This felt a\n bit heavy handed, but does perform significantly better for the default\n check_interval of 5 minutes (we\u0027re using a few seconds when testing for\n DRAM errors).  I could be convinced to go with this, if anyone felt it\n was not too aggressive.\n\nTesting:\n I used an error-injecting DIMM to create lots of correctable DRAM errors\n and verified that the polling interval accelerates.  The printk() only\n happens once per check_interval seconds.\n\nPatch:\n This patch is against 2.6.21-rc7.\n\nSigned-Off-By: Tim Hockin \u003cthockin@google.com\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "20280195f2a3d80c42a190959ca22108c93cd7e0",
      "tree": "5cfb384cf22c94f854a3a58acdb2b5341361e8bc",
      "parents": [
        "d824395c5994adbf7efe377cc67f732133270554"
      ],
      "author": {
        "name": "David Rientjes",
        "email": "rientjes@google.com",
        "time": "Wed May 02 19:27:09 2007 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Wed May 02 19:27:09 2007 +0200"
      },
      "message": "[PATCH] x86-64: fake numa for cpusets document\n\nCreate a document to explain how to use numa\u003dfake in conjunction with cpusets\nfor coarse memory resource management.\n\nAn attempt to get more awareness and testing for this feature.\n\nCc: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nCc: Paul Jackson \u003cpj@sgi.com\u003e\nCc: Christoph Lameter \u003cclameter@engr.sgi.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n"
    },
    {
      "commit": "382591d500bbcd20a44416c5e0e292708468587c",
      "tree": "deb996c4dca98cf36c432ddb710719f7d0202252",
      "parents": [
        "14694d736bb66d0ec250d05c81c6e98a19c229c6"
      ],
      "author": {
        "name": "David Rientjes",
        "email": "rientjes@google.com",
        "time": "Wed May 02 19:27:09 2007 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Wed May 02 19:27:09 2007 +0200"
      },
      "message": "[PATCH] x86-64: fixed size remaining fake nodes\n\nExtends the numa\u003dfake x86_64 command-line option to split the remaining system\nmemory into nodes of fixed size.  Any leftover memory is allocated to a final\nnode unless the command-line ends with a comma.\n\nFor example:\n  numa\u003dfake\u003d2*512,*128\tgives two 512M nodes and the remaining system\n\t\t\tmemory is split into nodes of 128M each.\n\nThis is beneficial for systems where the exact size of RAM is unknown or not\nnecessarily relevant, but the size of the remaining nodes to be allocated is\nknown based on their capacity for resource management.\n\nCc: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nCc: Paul Jackson \u003cpj@sgi.com\u003e\nCc: Christoph Lameter \u003cclameter@engr.sgi.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n"
    },
    {
      "commit": "14694d736bb66d0ec250d05c81c6e98a19c229c6",
      "tree": "76e752f310859c9904ee772068b2915b138e6572",
      "parents": [
        "8b8ca80e192b10eecc01fc44a2902510af86f73b"
      ],
      "author": {
        "name": "David Rientjes",
        "email": "rientjes@google.com",
        "time": "Wed May 02 19:27:09 2007 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Wed May 02 19:27:09 2007 +0200"
      },
      "message": "[PATCH] x86-64: split remaining fake nodes equally\n\nExtends the numa\u003dfake x86_64 command-line option to split the remaining\nsystem memory into equal-sized nodes.\n\nFor example:\nnuma\u003dfake\u003d2*512,4*\tgives two 512M nodes and the remaining system\n\t\t\tmemory is split into four approximately equal\n\t\t\tchunks.\n\nThis is beneficial for systems where the exact size of RAM is unknown or not\nnecessarily relevant, but the granularity with which nodes shall be allocated\nis known.\n\nCc: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nCc: Paul Jackson \u003cpj@sgi.com\u003e\nCc: Christoph Lameter \u003cclameter@engr.sgi.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n"
    },
    {
      "commit": "8b8ca80e192b10eecc01fc44a2902510af86f73b",
      "tree": "b5be871c62f4fd3f079a9ce7fcce0c09e87a9a5f",
      "parents": [
        "8280c0c58e9762a9fe29d550a9db81410de77691"
      ],
      "author": {
        "name": "David Rientjes",
        "email": "rientjes@google.com",
        "time": "Wed May 02 19:27:09 2007 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Wed May 02 19:27:09 2007 +0200"
      },
      "message": "[PATCH] x86-64: configurable fake numa node sizes\n\nExtends the numa\u003dfake x86_64 command-line option to allow for configurable\nnode sizes.  These nodes can be used in conjunction with cpusets for coarse\nmemory resource management.\n\nThe old command-line option is still supported:\n  numa\u003dfake\u003d32\tgives 32 fake NUMA nodes, ignoring the NUMA setup of the\n\t\tactual machine.\n\nBut now you may configure your system for the node sizes of your choice:\n  numa\u003dfake\u003d2*512,1024,2*256\n\t\tgives two 512M nodes, one 1024M node, two 256M nodes, and\n\t\tthe rest of system memory to a sixth node.\n\nThe existing hash function is maintained to support the various node sizes\nthat are possible with this implementation.\n\nEach node of the same size receives roughly the same amount of available\npages, regardless of any reserved memory with its address range.  The total\navailable pages on the system is calculated and divided by the number of equal\nnodes to allocate.  These nodes are then dynamically allocated and their\nborders extended until such time as their number of available pages reaches\nthe required size.\n\nConfigurable node sizes are recommended when used in conjunction with cpusets\nfor memory control because it eliminates the overhead associated with scanning\nthe zonelists of many smaller full nodes on page_alloc().\n\nCc: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: David Rientjes \u003crientjes@google.com\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nCc: Paul Jackson \u003cpj@sgi.com\u003e\nCc: Christoph Lameter \u003cclameter@engr.sgi.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\n"
    },
    {
      "commit": "9ce883becb83190061369940de9c415595836c9b",
      "tree": "61f2a1717e7d46751559a3dc897c05de8dfdb6d5",
      "parents": [
        "cf6387daf8858bdcb3e123034ca422e8979d73f1"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Tue Apr 24 13:05:37 2007 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Tue Apr 24 13:05:37 2007 +0200"
      },
      "message": "[PATCH] x86: Remove noreplacement option\n\nnoreplacement is dangerous on modern systems because it will not replace the\ncontext switch FNSAVE with SSE aware FXSAVE. But other places in the kernel still assume\nSSE and do FXSAVE and the CPU will then access FXSAVE information with\nFNSAVE and cause corruption.\n\nEasiest way to avoid this is to remove the option. It was mostly for paranoia\nreasons anyways and alternative()s have been stable for some time.\n\nThanks to Jeremy F. for reporting and helping debug it.\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "a98f0dd34d94ea0b5f3816196bea5dba467827bb",
      "tree": "019235e4d668b95366dd98dc6474716139c1584b",
      "parents": [
        "24ce0e96f2dea558762c994d054ea2f3c01fa95a"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Tue Feb 13 13:26:23 2007 +0100"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Tue Feb 13 13:26:23 2007 +0100"
      },
      "message": "[PATCH] x86-64: Allow to run a program when a machine check event is detected\n\nWhen a machine check event is detected (including a AMD RevF threshold\noverflow event) allow to run a \"trigger\" program. This allows user space\nto react to such events sooner.\n\nThe trigger is configured using a new trigger entry in the\nmachinecheck sysfs interface. It is currently shared between\nall CPUs.\n\nI also fixed the AMD threshold handler to run the machine\ncheck polling code immediately to actually log any events\nthat might have caused the threshold interrupt.\n\nAlso added some documentation for the mce sysfs interface.\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "57d307720c9a60038f134b0567ca302b88313a0a",
      "tree": "d84b4ca0181c3e8c638fd3c0c405bf61aa0f523e",
      "parents": [
        "44264261d8fb87849118e41b2735bd95db28126f"
      ],
      "author": {
        "name": "Randy Dunlap",
        "email": "randy.dunlap@oracle.com",
        "time": "Tue Feb 13 13:26:23 2007 +0100"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Tue Feb 13 13:26:23 2007 +0100"
      },
      "message": "[PATCH] x86-64: cleanup Doc/x86_64/ files\n\nFix typos.\nLots of whitespace changes for readability and consistency.\n\nSigned-off-by: Randy Dunlap \u003crandy.dunlap@oracle.com\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "5558870bfbcca10cfc7b13ab866687012ea3c9af",
      "tree": "45605429a4fd5eee9b05d5e7c852fa72bfd07a21",
      "parents": [
        "5809f9d442e9dbb23859e2c37d8c47043f6b5cc9"
      ],
      "author": {
        "name": "Karsten Weiss",
        "email": "K.Weiss@science-computing.de",
        "time": "Tue Feb 13 13:26:21 2007 +0100"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Tue Feb 13 13:26:21 2007 +0100"
      },
      "message": "[PATCH] x86-64: improved iommu documentation\n\n- add SWIOTLB config help text\n- mention Documentation/x86_64/boot-options.txt in\n  Documentation/kernel-parameters.txt\n- remove the duplication of the iommu kernel parameter documentation.\n- Better explanation of some of the iommu kernel parameter options.\n- \"32MB\u003c\u003corder\" instead of \"32MB^order\".\n- Mention the default \"order\" value.\n- list the four existing PCI-DMA mapping implementations of arch x86_64\n- group the iommu\u003d option keywords by PCI-DMA mapping implementation.\n- Distinguish iommu\u003d option keywords from number arguments.\n- Explain the meaning of DAC and SAC.\n\nSigned-off-by: Karsten Weiss \u003cknweiss@science-computing.de\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nAcked-by: Muli Ben-Yehuda \u003cmuli@il.ibm.com\u003e\nCc: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\n"
    },
    {
      "commit": "fea5f1e19611d94fbf3905875a427c4cb959cd06",
      "tree": "5a4c0584eea3ba74745e836e370fa1bbd0dacb02",
      "parents": [
        "bf81b46482c0fa8ea638e409d39768ea92a6b0f0"
      ],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@woody.osdl.org",
        "time": "Mon Jan 08 15:04:46 2007 -0800"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@woody.osdl.org",
        "time": "Mon Jan 08 15:04:46 2007 -0800"
      },
      "message": "Revert \"[PATCH] x86-64: Try multiple timer variants in check_timer\"\n\nThis reverts commit b026872601976f666bae77b609dc490d1834bf77, which has\nbeen linked to several problem reports with IO-APIC and the timer.\nMachines either don\u0027t boot because the timer doesn\u0027t happen, or we get\ndouble timer interrupts because we end up double-routing the timer irq\nthrough multiple interfaces.\n\nSee for example\n\n\thttp://lkml.org/lkml/2006/12/16/101\n\thttp://lkml.org/lkml/2007/1/3/9\n\thttp://bugzilla.kernel.org/show_bug.cgi?id\u003d7789\n\nabout some of the discussion.\n\nPatches to fix this cleanup exist (and have been confirmed to work fine\nat least for some of the affected cases) and we\u0027ll revisit it for\n2.6.21, but this late in the -rc series we\u0027re better off just reverting\nthe incomplete commit that caused the problems.\n\nSuggested-by: Adrian Bunk \u003cbunk@stusta.de\u003e\nCc: Eric W. Biederman \u003cebiederm@xmission.com\u003e\nCc: Yinghai Lu \u003cyinghai.lu@amd.com\u003e\nCc: Andrew Morton \u003cakpm@osdl.org\u003e\nCc: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "bff6547bb6a4e82c399d74e7fba78b12d2f162ed",
      "tree": "faeb4e95d243b0eb5cd6337739e4c14d154b282f",
      "parents": [
        "eae93755540bae18aff46b8a0e621b5d65bd5380"
      ],
      "author": {
        "name": "Muli Ben-Yehuda",
        "email": "muli@il.ibm.com",
        "time": "Thu Dec 07 02:14:07 2006 +0100"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Thu Dec 07 02:14:07 2006 +0100"
      },
      "message": "[PATCH] Calgary: allow compiling Calgary in but not using it by default\n\nThis patch makes it possible to compile Calgary in but not use it by\ndefault. In this mode, use \u0027iommu\u003dcalgary\u0027 to activate it.\n\nSigned-off-by: Muli Ben-Yehuda \u003cmuli@il.ibm.com\u003e\nSigned-off-by: Jon Mason \u003cjdmason@kudzu.us\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "b026872601976f666bae77b609dc490d1834bf77",
      "tree": "c1e9180ed6fd29cd025ba1d40407da8a3aeeaaa6",
      "parents": [
        "11a4180c0b03e2ee0c948fd8430ee092dc1625b3"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Thu Dec 07 02:14:06 2006 +0100"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Thu Dec 07 02:14:06 2006 +0100"
      },
      "message": "[PATCH] x86-64: Try multiple timer variants in check_timer\n\nInstead of adding all kinds of more quirks try various timer\nrouting variants in check_timer.\n\nIn particular this tries to handle quirks from:\n- Nvidia NF2-4 reference BIOS: wrong timer override\n- Asus: Wrong timer override but no HPET table\n- ATI: require timer disabled in 8259\n- Some boards: require timer enabled in 8259\n\nWe just try many of the the known variants in the hopefully right order\nin check_timer.\n\nTrying pin 0/2 on Nvidia suggested by Tim Hockin.\n\nTBD Experimental. Needs a lot of testing\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "d6bc8ac9e13e466e844313b590fbc49f7f1abdea",
      "tree": "c7c1ad2057d46d5e878ee034eaa9e4885a67d526",
      "parents": [
        "84eb8d0608af1576175307ed8fb3c8fde329e579"
      ],
      "author": {
        "name": "Matt LaPlante",
        "email": "kernel1@cyberdogtech.com",
        "time": "Tue Oct 03 22:54:15 2006 +0200"
      },
      "committer": {
        "name": "Adrian Bunk",
        "email": "bunk@stusta.de",
        "time": "Tue Oct 03 22:54:15 2006 +0200"
      },
      "message": "Fix typos in Documentation/: \u0027Q\u0027-\u0027R\u0027\n\nThis patch fixes typos in various Documentation txts. The patch addresses\nsome words starting with the letters \u0027Q\u0027-\u0027R\u0027.\n\nSigned-off-by: Matt LaPlante \u003ckernel1@cyberdogtech.com\u003e\nAcked-by: Randy Dunlap \u003crdunlap@xenotime.net\u003e\nSigned-off-by: Adrian Bunk \u003cbunk@stusta.de\u003e\n"
    },
    {
      "commit": "d802ab981da67d433fbb48f5640b7e3a9ba9fd54",
      "tree": "d0499267b945c506c5d3ad0c52608e0fd815385f",
      "parents": [
        "ded318ec80071557155604197ea45dcdb2a9ff2f"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Sat Sep 30 01:47:55 2006 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Sat Sep 30 01:47:55 2006 +0200"
      },
      "message": "[PATCH] Document iommu\u003dpanic\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "ece668401250c1d0419ed078c1d26ccdd30d3289",
      "tree": "6a8fc43edbee644cc6af3cf8252795c858a99b6a",
      "parents": [
        "b885808e185a4ec2dfe16c84434f79e95f0245b0"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Sat Sep 30 01:47:55 2006 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Sat Sep 30 01:47:55 2006 +0200"
      },
      "message": "[PATCH] Allow disabling DAC using command line options\n\nMight or might not work around some reported bugs on VIA systems.\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "352f7bae81faa2befa2a3c02b84478dce16b8fd6",
      "tree": "bcbaf7f7d1a71b2a7ded24e41a43b2ef848b8559",
      "parents": [
        "ba9c231f7499ff6918c069c72ff5fd836c76b963"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Tue Sep 26 10:52:31 2006 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Tue Sep 26 10:52:31 2006 +0200"
      },
      "message": "[PATCH] Add stack documentation document from Keith Owens\n\nDescribes the stack organization on x86-64.\n\nI changed it a bit and removed some obsolete information and the\nquestions.\n\nCc: kaos@sgi.com\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "538b5b419c7ae39a4b2deb15278da36102e42346",
      "tree": "cd85b52725c8ab1d024e6b6d9534e4ec611512ac",
      "parents": [
        "31679f38d88696ed032d59e457f1605c97e7d719"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Tue Sep 26 10:52:28 2006 +0200"
      },
      "committer": {
        "name": "Andi Kleen",
        "email": "andi@basil.nowhere.org",
        "time": "Tue Sep 26 10:52:28 2006 +0200"
      },
      "message": "[PATCH] Document backtracer selection options\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\n"
    },
    {
      "commit": "b783fd925cdd56d24d164e5bdcb072f2a67aedf4",
      "tree": "e2585d3a2f560eb5a313c92d0cd91df92358b1b6",
      "parents": [
        "b13761ecd1d9977d2083da243e051e9f29097aef"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Fri Jul 28 14:44:54 2006 +0200"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Fri Jul 28 19:28:00 2006 -0700"
      },
      "message": "[PATCH] x86_64: Document backtracer selection options\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "e465058d55a88feb4c7ecabe63eea7ea7147e206",
      "tree": "d431ed689e072415915694eecdfbcb9304287f01",
      "parents": [
        "0dc243ae10c8309c170a3af9f1adad1924a9f217"
      ],
      "author": {
        "name": "Jon Mason",
        "email": "jdmason@us.ibm.com",
        "time": "Mon Jun 26 13:58:14 2006 +0200"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Mon Jun 26 10:48:19 2006 -0700"
      },
      "message": "[PATCH] x86_64: Calgary IOMMU - Calgary specific bits\n\nThis patch hooks Calgary into the build, the x86-64 IOMMU\ninitialization paths, and introduces the Calgary specific bits.  The\nimplementation draws inspiration from both PPC (which has support for\nthe same chip but requires firmware support which we don\u0027t have on\nx86-64) and gart. Calgary is different from gart in that it support a\ntranslation table per PHB, as opposed to the single gart aperture.\n\nChanges from previous version:\n * Addition of boot-time disablement for bus-level translation/isolation\n   (e.g, enable userspace DMA for things like X)\n * Usage of newer IOMMU abstraction functions\n\nSigned-off-by: Muli Ben-Yehuda \u003cmuli@il.ibm.com\u003e\nSigned-off-by: Jon Mason \u003cjdmason@us.ibm.com\u003e\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "68a3a7feb08f960095072f28ec20f7900793c506",
      "tree": "1d458822fd4414997b6f12ad8a915f785d7e730f",
      "parents": [
        "9d99aaa31f5994d1923c3713ce9144c4c42332e1"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Fri Apr 07 19:49:18 2006 +0200"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Sun Apr 09 11:53:16 2006 -0700"
      },
      "message": "[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64\n\nFrom: Keith Mannthey, Andi Kleen\n\nImplement memory hotadd without sparsemem. The memory in the SRAT\nhotadd area is just preserved instead and can be activated later.\n\nThere are a few restrictions:\n- Only one continuous hotadd area allowed per node\n\nThe main problem is dealing with the many buggy SRAT tables\nthat are out there. The strategy here is to reject anything\nsuspicious.\n\nOriginally from Keith Mannthey, with several hacks and changes by AK\nand also contributions from Andrew Morton\n\n[ TBD: Problems pointed out by KAMEZAWA Hiroyuki \u003ckamezawa.hiroyu@jp.fujitsu.com\u003e:\n\n 1) Goto\u0027s rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG\u003dn.\n\n    Rebuilding zonelist is necessary when the system has just memory \u003c\n    4G at boot, and hot add memory \u003e 4G.  because x86_64 has DMA32,\n    ZONE_NORAML is not included into zonelist at boot time if system\n    doesn\u0027t have memory \u003e4G at boot.\n\n    [AK: should just force the higher zones at boot time when SRAT tells us]\n\n 2) zone and node\u0027s spanned_pages and present_pages are not incremented.\n    They should be.\n\n    For example, our server (ia64/Fujitsu PrimeQuest) can equip memory\n    from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have\n    possible 1T +memory.  (Microsoft requires \"write all possible memory\n    in SRAT\") When we reserve memmap for possible 1T memory, Linux will\n    not work well in +minimum 4G configuraion ;)\n\n    [AK: needs limiting to 5-10% of max memory]\n ]\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "ab9b32ee626e9b6df4ce2560a70ae15e62423cf4",
      "tree": "4bae7b0a38b9e69d33dc7b80ad434274019d9e38",
      "parents": [
        "e8b917775b572bc27de105f1317c2de4335db5b3"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Sun Feb 26 04:18:52 2006 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Sun Feb 26 09:53:31 2006 -0800"
      },
      "message": "[PATCH] x86_64: Better ATI timer fix\n\nThe previous experiment for using apicmaintimer on ATI systems didn\u0027t\nwork out very well.  In particular laptops with C2/C3 support often\ndon\u0027t let it tick during idle, which makes it useless.  There were also\nsome other bugs that made the apicmaintimer often not used at all.\n\nI tried some other experiments - running timer over RTC and some other\nthings but they didn\u0027t really work well neither.\n\nI rechecked the specs now and it turns out this simple change is\nactually enough to avoid the double ticks on the ATI systems.  We just\nturn off IRQ 0 in the 8254 and only route it directly using the IO-APIC.\n\nI tested it on a few ATI systems and it worked there.  In fact it worked\non all chipsets (NVidia, Intel, AMD, ATI) I tried it on.\n\nAccording to the ACPI spec routing should always work through the\nIO-APIC so I think it\u0027s the correct thing to do anyways (and most of the\nold gunk in check_timer should be thrown away for x86-64).\n\nBut for 2.6.16 it\u0027s best to do a fairly minimal change:\n - Use the known to be working everywhere-but-ATI IRQ0 both over 8254\n   and IO-APIC setup everywhere\n - Except on ATI disable IRQ0 in the 8254\n - Remove the code to select apicmaintimer on ATI chipsets\n - Add some boot options to allow to override this (just paranoia)\n\nIn 2.6.17 I hope to switch the default over to this for everybody.\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "0c3749c41f5eee0da36bbf92b2793338b4d8574f",
      "tree": "3af004c5533ab04297004f5ec40da4c6801b8fda",
      "parents": [
        "099f318b8d97490e58c595eb1b6d3415da5ccc03"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Fri Feb 03 21:51:41 2006 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Sat Feb 04 16:43:15 2006 -0800"
      },
      "message": "[PATCH] x86_64: Calibrate APIC timer using PM timer\n\nOn some broken motherboards (at least one NForce3 based AMD64 laptop)\nthe PIT timer runs at a incorrect frequency.  This patch adds a new\noption \"apicpmtimer\" that allows to use the APIC timer and calibrate it\nusing the PMTimer.  It requires the earlier patch that allows to run the\nmain timer from the APIC.\n\nSpecifying apicpmtimer implies apicmaintimer.\n\nThe option defaults to off for now.\n\nI tested it on a few systems and the resulting APIC timer frequencies\nwere usually a bit off, but always \u003c1%, which should be tolerable.\n\nTBD figure out heuristic to enable this automatically on the affected\nsystems TBD perhaps do it on all NForce3s or using DMI?\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "73dea47faeb96d54a984b9d7f4de564816966354",
      "tree": "6b1f090e9071e281367bc106c1b2ab05cb43e459",
      "parents": [
        "76b461c21468f41837283b7888d55f1c0671f719"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Fri Feb 03 21:50:50 2006 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Sat Feb 04 16:43:13 2006 -0800"
      },
      "message": "[PATCH] x86_64: Allow to run main time keeping from the local APIC interrupt\n\nAnother piece from the no-idle-tick patch.\n\nThis can be enabled with the \"apicmaintimer\" option.\n\nThis is mainly useful when the PIT/HPET interrupt is unreliable.\nNote there are some systems that are known to stop the APIC\ntimer in C3. For those it will never work, but this case\nshould be automatically detected.\n\nIt also only works with PM timer right now. When HPET is used\nthe way the main timer handler computes the delay doesn\u0027t work.\n\nIt should be a bit more efficient because there is one less\nregular interrupt to process on the boot processor.\n\nRequires earlier bugfix from Venkatesh\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "575c96871832e91983248aebb02fc072e5fd17c9",
      "tree": "2de2ce229506f1e0e866ef00c9cf7e43489e295b",
      "parents": [
        "a8e82ef97a1d1227575698935ec86ae4d02c372c"
      ],
      "author": {
        "name": "Adrian Bunk",
        "email": "bunk@stusta.de",
        "time": "Sun Jan 15 02:00:17 2006 +0100"
      },
      "committer": {
        "name": "Adrian Bunk",
        "email": "bunk@r063144.stusta.swh.mhn.de",
        "time": "Sun Jan 15 02:00:17 2006 +0100"
      },
      "message": "spelling: s/appropiate/appropriate/\n\nSigned-off-by: Adrian Bunk \u003cbunk@stusta.de\u003e\n"
    },
    {
      "commit": "7eb903f4a5c35c8310f0aa7b0e94aae0b826d837",
      "tree": "39489c4edd583ae398d7987019a4ba7315bdfcee",
      "parents": [
        "f62a91f6911479642c0018290d4248ace4287648"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Wed Jan 11 22:42:39 2006 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Wed Jan 11 19:01:11 2006 -0800"
      },
      "message": "[PATCH] x86_64: Add documentation for CPU hotplug ACPI extension\n\nCc: len.brown@intel.com, ashok.ray@intel.com\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "f62a91f6911479642c0018290d4248ace4287648",
      "tree": "f5d941d1f1dab0bf462795c47fa8b616d31a833f",
      "parents": [
        "92934bcbf96bc9dc931c40ca5f1a57685b7b813b"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Wed Jan 11 22:42:35 2006 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Wed Jan 11 19:01:11 2006 -0800"
      },
      "message": "[PATCH] x86_64: Don\u0027t reserve hotplug CPUs by default\n\nMost users don\u0027t need it so no need to waste memory.\nThis means an user has to specify the appropiate number of\nhotplug CPUs on the command line with additional_cpus\u003d...\nor fix their BIOS to follow the convention in\nDocumentation/x86-64/cpu-hotplug-spec\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "9e43e1b7c7c9872da032442d8e4bb112a02d16f4",
      "tree": "04399c723a8fbd3f859a462add5db34c389ebf45",
      "parents": [
        "ffd10a2b77bca50dd05ba26acd5a6e68bcc8f61f"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Sat Nov 05 17:25:54 2005 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Mon Nov 14 19:55:17 2005 -0800"
      },
      "message": "[PATCH] x86_64: Remove CONFIG_CHECKING and add command line option for pagefault tracing\n\nCONFIG_CHECKING covered some debugging code used in the early times\nof the port. But it wasn\u0027t even SMP safe for quite some time\nand the bugs it checked for seem to be gone.\n\nThis patch removes all the code to verify GS at kernel entry. There\nhaven\u0027t been any new bugs in this area for a long time.\n\nPreviously it also covered the sysctl for the page fault tracing.\nThat didn\u0027t make much sense because that code was unconditionally\ncompiled in. I made that a boot option now because it is typically\nonly useful at boot.\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "e583538f077d5f70191670b47a046ba436ec3428",
      "tree": "576298f190ce8a4f321bb84efce2ace41aa12194",
      "parents": [
        "b0bd35e622ffbda2c01dc67a0381c6a18817a29a"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Sat Nov 05 17:25:54 2005 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Mon Nov 14 19:55:17 2005 -0800"
      },
      "message": "[PATCH] x86_64: Log machine checks from boot on Intel systems\n\nThe logging for boot errors was turned off because it was broken\non some AMD systems. But give Intel EM64T systems a chance because they are\nsupposed to be correct there.\n\nThe advantage is that there is a chance to actually log uncorrected\nmachine checks after the reset.\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "420f8f68c9c5148dddf946bebdbc7eacde2172cb",
      "tree": "59cf49ac38e0c7560bb57d988869f67b1ff20342",
      "parents": [
        "485832a5d928facd82f1525270d9f048da2063a1"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Sat Nov 05 17:25:54 2005 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Mon Nov 14 19:55:15 2005 -0800"
      },
      "message": "[PATCH] x86_64: New heuristics to find out hotpluggable CPUs.\n\nWith a NR_CPUS\u003d\u003d128 kernel with CPU hotplug enabled we would waste 4MB\non per CPU data of all possible CPUs.  The reason was that HOTPLUG\nalways set up possible map to NR_CPUS cpus and then we need to allocate\nthat much (each per CPU data is roughly ~32k now)\n\nThe underlying problem is that ACPI didn\u0027t tell us how many hotplug CPUs\nthe platform supports.  So the old code just assumed all, which would\nlead to this memory wastage.\n\nThis implements some new heuristics:\n\n - If the BIOS specified disabled CPUs in the ACPI/mptables assume they\n   can be enabled later (this is bending the ACPI specification a bit,\n   but seems like a obvious extension)\n - The user can overwrite it with a new additionals_cpus\u003dNUM option\n - Otherwise use half of the available CPUs or 2, whatever is more.\n\nCc: ashok.raj@intel.com\nCc: len.brown@intel.com\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "8315eca25583c369e28f48909d3341dc21d6214d",
      "tree": "271677ef4bb1159284c0920aabc903218868fc1f",
      "parents": [
        "59170891049cc469777a392e0b3f5aa7aad784a6"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Sat Nov 05 17:25:54 2005 +0100"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Mon Nov 14 19:55:15 2005 -0800"
      },
      "message": "[PATCH] x86_64: Some clarifications for Documention/x86_64/mm.txt\n\nI got some questions on this, so just fix up the documentation.\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "8c566ef5f3eb5d9daf61d0fa31d82c211f601ba0",
      "tree": "646730e657f27a23b3081fafafe494121e0d8075",
      "parents": [
        "673242c10d535bfe238d9d8e82ac93432d35b88e"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Mon Sep 12 18:49:24 2005 +0200"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Mon Sep 12 10:49:57 2005 -0700"
      },
      "message": "[PATCH] x86-64: Add command line option to set machine check tolerance level\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "d5172f263f76ca6d588f533c0989b22521fcbac2",
      "tree": "a4a017f686f101b5817f6c1abbc84518335d497e",
      "parents": [
        "cf7bee5a0bf270a4eace0be39329d6ac0136cc47"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@muc.de",
        "time": "Sun Aug 07 09:42:07 2005 -0700"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Sun Aug 07 10:00:37 2005 -0700"
      },
      "message": "[PATCH] x86_64: ignore machine checks from boot time\n\nDon\u0027t log machine check events left over from boot.  Too many BIOSes leave\nbogus events in there.\n\nThis unfortunately also makes it impossible to log events that caused a\nreboot.  For people with non broken BIOS there is mce\u003dbootlog\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "ef4d7cbea773a77b36e732779cab4018ba2c037b",
      "tree": "c9aeb9e716ee2d83204724d6e2baa2b2e81caf0e",
      "parents": [
        "a54649b801e5fe69755dc55e6589f7a65af25d79"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Thu Jul 28 21:15:34 2005 -0700"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@g5.osdl.org",
        "time": "Thu Jul 28 21:45:59 2005 -0700"
      },
      "message": "[PATCH] x86_64: Some updates for boot-options.txt\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "14d98cad82b78956957e95567b8b5fb38ec5859f",
      "tree": "ab1b15e3a2524805495d21be074551f9e553e270",
      "parents": [
        "607a16858397829806c5a4db999ce6daf327f98c"
      ],
      "author": {
        "name": "Andi Kleen",
        "email": "ak@suse.de",
        "time": "Fri May 20 14:27:59 2005 -0700"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@ppc970.osdl.org",
        "time": "Fri May 20 15:48:21 2005 -0700"
      },
      "message": "[PATCH] x86_64: Add option to disable timer check\n\nThis works around the too fast timer seen on some ATI boards.\n\nI don\u0027t feel confident enough about it yet to enable it by default, but give\nusers the option.\n\nPatch and debugging from Christopher Allen Wing \u003cwingc@engin.umich.edu\u003e, with\nminor tweaks (renamed the option and documented it)\n\nSigned-off-by: Andi Kleen \u003cak@suse.de\u003e\nSigned-off-by: Andrew Morton \u003cakpm@osdl.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@osdl.org\u003e\n"
    },
    {
      "commit": "1da177e4c3f41524e886b7f1b8a0c1fc7321cac2",
      "tree": "0bba044c4ce775e45a88a51686b5d9f90697ea9d",
      "parents": [],
      "author": {
        "name": "Linus Torvalds",
        "email": "torvalds@ppc970.osdl.org",
        "time": "Sat Apr 16 15:20:36 2005 -0700"
      },
      "committer": {
        "name": "Linus Torvalds",
        "email": "torvalds@ppc970.osdl.org",
        "time": "Sat Apr 16 15:20:36 2005 -0700"
      },
      "message": "Linux-2.6.12-rc2\n\nInitial git repository build. I\u0027m not bothering with the full history,\neven though we have it. We can create a separate \"historical\" git\narchive of that later if we want to, and in the meantime it\u0027s about\n3.2GB when imported into git - space that would just make the early\ngit days unnecessarily complicated, when we don\u0027t have a lot of good\ninfrastructure for it.\n\nLet it rip!\n"
    }
  ]
}