)]}' { "log": [ { "commit": "5f0b819995e172f48fdcd91335a2126ba7d9deae", "tree": "8e83ba4c380ce0e546fb8702e7d688f180cb4646", "parents": [ "844a5fe219cf472060315971e15cbf97674a3324" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Wed Mar 09 14:28:02 2016 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 10 11:26:10 2016 +0100" }, "message": "KVM: MMU: fix reserved bit check for ept\u003d0/CR0.WP\u003d0/CR4.SMEP\u003d1/EFER.NX\u003d0\n\nKVM has special logic to handle pages with pte.u\u003d1 and pte.w\u003d0 when\nCR0.WP\u003d1. These pages\u0027 SPTEs flip continuously between two states:\nU\u003d1/W\u003d0 (user and supervisor reads allowed, supervisor writes not allowed)\nand U\u003d0/W\u003d1 (supervisor reads and writes allowed, user writes not allowed).\n\nWhen SMEP is in effect, however, U\u003d0 will enable kernel execution of\nthis page. To avoid this, KVM also sets NX\u003d1 in the shadow PTE together\nwith U\u003d0, making the two states U\u003d1/W\u003d0/NX\u003dgpte.NX and U\u003d0/W\u003d1/NX\u003d1.\nWhen guest EFER has the NX bit cleared, the reserved bit check thinks\nthat the latter state is invalid; teach it that the smep_andnot_wp case\nwill also use the NX bit of SPTEs.\n\nCc: stable@vger.kernel.org\nReviewed-by: Xiao Guangrong \u003cguangrong.xiao@linux.inel.com\u003e\nFixes: c258b62b264fdc469b6d3610a907708068145e3b\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "844a5fe219cf472060315971e15cbf97674a3324", "tree": "ffc7eb5a0cd87af357eec478e12c9dc902c3223f", "parents": [ "313f636d5c490c9741d3f750dc8da33029edbc6b" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Tue Mar 08 12:13:39 2016 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Thu Mar 10 11:26:07 2016 +0100" }, "message": "KVM: MMU: fix ept\u003d0/pte.u\u003d1/pte.w\u003d0/CR0.WP\u003d0/CR4.SMEP\u003d1/EFER.NX\u003d0 combo\n\nYes, all of these are needed. :) This is admittedly a bit odd, but\nkvm-unit-tests access.flat tests this if you run it with \"-cpu host\"\nand of course ept\u003d0.\n\nKVM runs the guest with CR0.WP\u003d1, so it must handle supervisor writes\nspecially when pte.u\u003d1/pte.w\u003d0/CR0.WP\u003d0. Such writes cause a fault\nwhen U\u003d1 and W\u003d0 in the SPTE, but they must succeed because CR0.WP\u003d0.\nWhen KVM gets the fault, it sets U\u003d0 and W\u003d1 in the shadow PTE and\nrestarts execution. This will still cause a user write to fault, while\nsupervisor writes will succeed. User reads will fault spuriously now,\nand KVM will then flip U and W again in the SPTE (U\u003d1, W\u003d0). User reads\nwill be enabled and supervisor writes disabled, going back to the\noriginary situation where supervisor writes fault spuriously.\n\nWhen SMEP is in effect, however, U\u003d0 will enable kernel execution of\nthis page. To avoid this, KVM also sets NX\u003d1 in the shadow PTE together\nwith U\u003d0. If the guest has not enabled NX, the result is a continuous\nstream of page faults due to the NX bit being reserved.\n\nThe fix is to force EFER.NX\u003d1 even if the CPU is taking care of the EFER\nswitch. (All machines with SMEP have the CPU_LOAD_IA32_EFER vm-entry\ncontrol, so they do not use user-return notifiers for EFER---if they did,\nEFER.NX would be forced to the same value as the host).\n\nThere is another bug in the reserved bit check, which I\u0027ve split to a\nseparate patch for easier application to stable kernels.\n\nCc: stable@vger.kernel.org\nCc: Andy Lutomirski \u003cluto@amacapital.net\u003e\nReviewed-by: Xiao Guangrong \u003cguangrong.xiao@linux.intel.com\u003e\nFixes: f6577a5fa15d82217ca73c74cd2dcbc0f6c781dd\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "313f636d5c490c9741d3f750dc8da33029edbc6b", "tree": "56e9b7133f98991b710c8776712a2e574c41b78c", "parents": [ "9522b37f5a8c7bfabe46eecadf2e130f1103f337" ], "author": { "name": "David Matlack", "email": "dmatlack@google.com", "time": "Tue Mar 08 16:19:44 2016 -0800" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Wed Mar 09 11:54:14 2016 +0100" }, "message": "kvm: cap halt polling at exactly halt_poll_ns\n\nWhen growing halt-polling, there is no check that the poll time exceeds\nthe limit. It\u0027s possible for vcpu-\u003ehalt_poll_ns grow once past\nhalt_poll_ns, and stay there until a halt which takes longer than\nvcpu-\u003ehalt_poll_ns. For example, booting a Linux guest with\nhalt_poll_ns\u003d11000:\n\n ... kvm:kvm_halt_poll_ns: vcpu 0: halt_poll_ns 0 (shrink 10000)\n ... kvm:kvm_halt_poll_ns: vcpu 0: halt_poll_ns 10000 (grow 0)\n ... kvm:kvm_halt_poll_ns: vcpu 0: halt_poll_ns 20000 (grow 10000)\n\nSigned-off-by: David Matlack \u003cdmatlack@google.com\u003e\nFixes: aca6ff29c4063a8d467cdee241e6b3bf7dc4a171\nCc: stable@vger.kernel.org\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "9522b37f5a8c7bfabe46eecadf2e130f1103f337", "tree": "8b8b61f1daa3b5f1732f2d10bc1ce66e5bf62a2f", "parents": [ "8bb9b9ccff5b7ce5084b5111244a92fd1d0daf16" ], "author": { "name": "David Hildenbrand", "email": "dahi@linux.vnet.ibm.com", "time": "Tue Mar 08 12:24:30 2016 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Tue Mar 08 12:47:01 2016 +0100" }, "message": "KVM: s390: correct fprs on SIGP (STOP AND) STORE STATUS\n\nWith MACHINE_HAS_VX, we convert the floating point registers from the\nvector registeres when storing the status. For other VCPUs, these are\nstored to vcpu-\u003erun-\u003es.regs.vrs, but we are using current-\u003ethread.fpu.vxrs,\nwhich resolves to the currently loaded VCPU.\n\nSo kvm_s390_store_status_unloaded() currently writes the wrong floating\npoint registers (converted from the vector registers) when called from\nanother VCPU on a z13.\n\nThis is only the case for old user space not handling SIGP STORE STATUS and\nSIGP STOP AND STORE STATUS, but relying on the kernel implementation. All\nother calls come from the loaded VCPU via kvm_s390_store_status().\n\nFixes: 9abc2a08a7d6 (KVM: s390: fix memory overwrites when vx is disabled)\nReviewed-by: Christian Borntraeger \u003cborntraeger@de.ibm.com\u003e\nCc: stable@vger.kernel.org # v4.4+\nSigned-off-by: David Hildenbrand \u003cdahi@linux.vnet.ibm.com\u003e\nSigned-off-by: Christian Borntraeger \u003cborntraeger@de.ibm.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "8bb9b9ccff5b7ce5084b5111244a92fd1d0daf16", "tree": "c4d42efe0fdde7dc07e469939bc1f17c7886f213", "parents": [ "7099e2e1f4d9051f31bbfa5803adf954bb5d76ef", "ccec44563b18a0ce90e2d4f332784b3cb25c8e9c" ], "author": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Tue Mar 08 12:46:50 2016 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Tue Mar 08 12:46:50 2016 +0100" }, "message": "Merge branch \u0027kvm-ppc-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc into HEAD\n" }, { "commit": "7099e2e1f4d9051f31bbfa5803adf954bb5d76ef", "tree": "229093e55b76aab1b2f97bb3ef8891431d47fdb9", "parents": [ "f6cede5b49e822ebc41a099fe41ab4989f64e2cb" ], "author": { "name": "Radim Krčmář", "email": "rkrcmar@redhat.com", "time": "Fri Mar 04 15:08:42 2016 +0100" }, "committer": { "name": "Paolo Bonzini", "email": "pbonzini@redhat.com", "time": "Tue Mar 08 12:46:46 2016 +0100" }, "message": "KVM: VMX: disable PEBS before a guest entry\n\nLinux guests on Haswell (and also SandyBridge and Broadwell, at least)\nwould crash if you decided to run a host command that uses PEBS, like\n perf record -e \u0027cpu/mem-stores/pp\u0027 -a\n\nThis happens because KVM is using VMX MSR switching to disable PEBS, but\nSDM [2015-12] 18.4.4.4 Re-configuring PEBS Facilities explains why it\nisn\u0027t safe:\n When software needs to reconfigure PEBS facilities, it should allow a\n quiescent period between stopping the prior event counting and setting\n up a new PEBS event. The quiescent period is to allow any latent\n residual PEBS records to complete its capture at their previously\n specified buffer address (provided by IA32_DS_AREA).\n\nThere might not be a quiescent period after the MSR switch, so a CPU\nends up using host\u0027s MSR_IA32_DS_AREA to access an area in guest\u0027s\nmemory. (Or MSR switching is just buggy on some models.)\n\nThe guest can learn something about the host this way:\nIf the guest doesn\u0027t map address pointed by MSR_IA32_DS_AREA, it results\nin #PF where we leak host\u0027s MSR_IA32_DS_AREA through CR2.\n\nAfter that, a malicious guest can map and configure memory where\nMSR_IA32_DS_AREA is pointing and can therefore get an output from\nhost\u0027s tracing.\n\nThis is not a critical leak as the host must initiate with PEBS tracing\nand I have not been able to get a record from more than one instruction\nbefore vmentry in vmx_vcpu_run() (that place has most registers already\noverwritten with guest\u0027s).\n\nWe could disable PEBS just few instructions before vmentry, but\ndisabling it earlier shouldn\u0027t affect host tracing too much.\nWe also don\u0027t need to switch MSR_IA32_PEBS_ENABLE on VMENTRY, but that\noptimization isn\u0027t worth its code, IMO.\n\n(If you are implementing PEBS for guests, be sure to handle the case\n where both host and guest enable PEBS, because this patch doesn\u0027t.)\n\nFixes: 26a4f3c08de4 (\"perf/x86: disable PEBS on a guest entry.\")\nCc: \u003cstable@vger.kernel.org\u003e\nReported-by: Jiří Olša \u003cjolsa@redhat.com\u003e\nSigned-off-by: Radim Krčmář \u003crkrcmar@redhat.com\u003e\nSigned-off-by: Paolo Bonzini \u003cpbonzini@redhat.com\u003e\n" }, { "commit": "ccec44563b18a0ce90e2d4f332784b3cb25c8e9c", "tree": "00879794b39f4561398926e0cbc6ba9588827bc3", "parents": [ "2680d6da455b636dd006636780c0f235c6561d70" ], "author": { "name": "Paul Mackerras", "email": "paulus@samba.org", "time": "Sat Mar 05 19:34:39 2016 +1100" }, "committer": { "name": "Paul Mackerras", "email": "paulus@samba.org", "time": "Tue Mar 08 13:36:42 2016 +1100" }, "message": "KVM: PPC: Book3S HV: Sanitize special-purpose register values on guest exit\n\nThomas Huth discovered that a guest could cause a hard hang of a\nhost CPU by setting the Instruction Authority Mask Register (IAMR)\nto a suitable value. It turns out that this is because when the\ncode was added to context-switch the new special-purpose registers\n(SPRs) that were added in POWER8, we forgot to add code to ensure\nthat they were restored to a sane value on guest exit.\n\nThis adds code to set those registers where a bad value could\ncompromise the execution of the host kernel to a suitable neutral\nvalue on guest exit.\n\nCc: stable@vger.kernel.org # v3.14+\nFixes: b005255e12a3\nReported-by: Thomas Huth \u003cthuth@redhat.com\u003e\nReviewed-by: David Gibson \u003cdavid@gibson.dropbear.id.au\u003e\nSigned-off-by: Paul Mackerras \u003cpaulus@samba.org\u003e\n" }, { "commit": "f6cede5b49e822ebc41a099fe41ab4989f64e2cb", "tree": "6c7987a61dbb25fe192becc6515d95dd5980d0a1", "parents": [ "d55e08c8448def74ae4236cd6e1c3714d78f5abd" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 14:48:03 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 14:48:03 2016 -0800" }, "message": "Linux 4.5-rc7\n" }, { "commit": "d55e08c8448def74ae4236cd6e1c3714d78f5abd", "tree": "fc8fc4e1c64450922a68a53c09255bde50b10515", "parents": [ "dd273a8071124c2235c45d2ca756acc90ca31ed3", "f3c87e99f401a0ff0ef941346ec6e790f0fd526e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 14:14:54 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 14:14:54 2016 -0800" }, "message": "Merge tag \u0027armsoc-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc\n\nPull ARM SoC fix from Olof Johansson:\n \"Tiny fixes branch this week, in fact only one patch.\n\n Turns out the USB support for a Renesas board was developed on a\n pre-release board that ended up being changed before shipping. To\n avoid breakage on those boards, and avoid confusion, it\u0027s a reasonable\n idea to patch now instead of later. There are no known users of the\n pre-release variant any more\"\n\n* tag \u0027armsoc-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:\n ARM: dts: porter: remove enable prop from HS-USB device node\n" }, { "commit": "dd273a8071124c2235c45d2ca756acc90ca31ed3", "tree": "805bb65c17779ce1f16245b56d5e9ccbd7309f27", "parents": [ "21b27a74ec0f4e52e001912c89570a49494705a2", "f474c8c857d996f34c39f66bbed23faaa739fad6" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 13:51:27 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 13:51:27 2016 -0800" }, "message": "Merge branch \u0027fixes\u0027 of git://ftp.arm.linux.org.uk/~rmk/linux-arm\n\nPull ARM fixes from Russell King:\n \"Just two ARM fixes this time: one to fix the hyp-stub for older ARM\n CPUs, and another to fix the set_memory_xx() permission functions to\n deal with zero sizes correctly\"\n\n* \u0027fixes\u0027 of git://ftp.arm.linux.org.uk/~rmk/linux-arm:\n ARM: 8544/1: set_memory_xx fixes\n ARM: 8534/1: virt: fix hyp-stub build for pre-ARMv7 CPUs\n" }, { "commit": "21b27a74ec0f4e52e001912c89570a49494705a2", "tree": "c9f1bc13a14ad1bd41739e12b92d9510d5b6e6e5", "parents": [ "a58b9adaf9f0913af2e5bf69565266335033b4fd", "5ea5c5e0a7f70b256417d3b6e36bd9851504babd" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:31:13 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:31:13 2016 -0800" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client\n\nPull ceph fix from Sage Weil:\n \"This is a final commit we missed to align the protocol compatibility\n with the feature bits.\n\n It decodes a few extra fields in two different messages and reports\n EIO when they are used (not yet supported)\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:\n ceph: initial CEPH_FEATURE_FS_FILE_LAYOUT_V2 support\n" }, { "commit": "a58b9adaf9f0913af2e5bf69565266335033b4fd", "tree": "45f91121352f96528bb02c4e26920405d6f3013d", "parents": [ "1306b0471ff11409e4f18f3cde00215d367e007a", "e4f6daac20332448529b11f09388f1d55ef2084c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:24:05 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:24:05 2016 -0800" }, "message": "Merge tag \u0027upstream-4.5-rc7\u0027 of git://git.infradead.org/linux-ubifs\n\nPull UBI fix from Richard Weinberger:\n \"This contains a single bug fix for UBI\"\n\n* tag \u0027upstream-4.5-rc7\u0027 of git://git.infradead.org/linux-ubifs:\n ubi: Fix out of bounds write in volume update code\n" }, { "commit": "1306b0471ff11409e4f18f3cde00215d367e007a", "tree": "ee041c404f964929b62fc734fa65b7afea58643f", "parents": [ "76d9c6c1c6eb115c2c686d5cbf10a68857511853", "ad32a1f3c36b046b2e7b0361b88c18e3af61419e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:19:28 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:19:28 2016 -0800" }, "message": "Merge branch \u0027for-linus-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rw/uml\n\nPull UML fixes from Richard Weinberger:\n \"This contains three bug/build fixes\"\n\n* \u0027for-linus-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rw/uml:\n um: use %lx format specifiers for unsigned longs\n um: Export pm_power_off\n Revert \"um: Fix get_signal() usage\"\n" }, { "commit": "76d9c6c1c6eb115c2c686d5cbf10a68857511853", "tree": "89b6e8b1de5c6018922be1ddf3e217908d357023", "parents": [ "b8155fe1b2963ab2c53140929eb2362afcb9ccfe", "e723e3f7f9591b79e8c56b3d7c5a204a9c571b55" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:14:16 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:14:16 2016 -0800" }, "message": "Merge branch \u0027upstream\u0027 of git://git.linux-mips.org/pub/scm/ralf/upstream-linus\n\nPull MIPS fixes from Ralf Baechle:\n \"Another round of fixes for 4.5:\n\n - Fix the use of an undocumented syntactial variant of the .type\n pseudo op which is not supported by the LLVM assembler.\n - Fix invalid initialization on S-cache-less systems.\n - Fix possible information leak from the kernel stack for SIGFPE.\n - Fix handling of copy_{from,to}_user() return value in KVM\n - Fix the last instance of irq_to_gpio() which now was causing build\n errors\"\n\n* \u0027upstream\u0027 of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:\n MIPS: traps: Fix SIGFPE information leak from `do_ov\u0027 and `do_trap_or_bp\u0027\n MIPS: kvm: Fix ioctl error handling.\n MIPS: scache: Fix scache init with invalid line size.\n MIPS: Avoid variant of .type unsupported by LLVM Assembler\n MIPS: jz4740: Fix surviving instance of irq_to_gpio()\n" }, { "commit": "b8155fe1b2963ab2c53140929eb2362afcb9ccfe", "tree": "5e63f57ce23f68094676205059bb53df4b7b5ec3", "parents": [ "7addb7fa1085e29398c2aef43f0b98c3c4b7e749", "37c5e942bb2eedd98c1cd1fa1f94d79f6b830c38" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:08:06 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:08:06 2016 -0800" }, "message": "Merge tag \u0027powerpc-4.5-5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux\n\nPull powerpc fixes from Michael Ellerman:\n - cxl: Fix PSL timebase synchronization detection from Frederic Barrat\n - Fix oops when destroying hw_breakpoint event from Ravi Bangoria\n - Avoid lbarx on e5500 from Scott Wood\n\n* tag \u0027powerpc-4.5-5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:\n powerpc/fsl-book3e: Avoid lbarx on e5500\n powerpc/hw_breakpoint: Fix oops when destroying hw_breakpoint event\n cxl: Fix PSL timebase synchronization detection\n" }, { "commit": "7addb7fa1085e29398c2aef43f0b98c3c4b7e749", "tree": "018732a4fc23772498f64e62706066087fc733a4", "parents": [ "3e09b7e93dae6cd9fcabc5dd36fc6973d3049c30", "7314d22a2f5bd40468d57768be368c3d9b4bd726" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:03:34 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 11:03:34 2016 -0800" }, "message": "Merge branch \u0027i2c/for-current\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux\n\nPull i2c fix from Wolfram Sang:\n \"One I2C bugfix ensuring correct memory allocation in a driver\"\n\n* \u0027i2c/for-current\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:\n i2c: brcmstb: allocate correct amount of memory for regmap\n" }, { "commit": "3e09b7e93dae6cd9fcabc5dd36fc6973d3049c30", "tree": "1014554302d56cdb13b8ce6a2c2e3615baad2d97", "parents": [ "67944024c1cdd897e49a09b0d6af3ea38d1388ca", "861c3849222b7cadfd0da63f9c7573c2f15ba535" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 10:50:00 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Mar 06 10:50:00 2016 -0800" }, "message": "Merge tag \u0027usb-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb\n\nPull USB fixes from Greg KH:\n \"Here are some USB driver ids for 4.5-rc7, and the removal of a driver\n we merged in 4.5-rc1 but it turns out it\u0027s not needed as the hardware\n is the same as a driver we already have in the tree.\n\n This was only figured out after doing a lot of cleanup on it, gotta\n love vendor-provided drivers... The new device ids for the devices\n for this driver will be added later on when testing is completed, but\n for now, we will remove the driver to keep people from accidentally\n cleaning it up.\n\n All of these have been in linux-next for a while with no reported\n issues\"\n\n* tag \u0027usb-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:\n USB: qcserial: add Sierra Wireless EM74xx device ID\n Revert \"USB: serial: add Moxa UPORT 11x0 driver\"\n USB: serial: option: add support for Quectel UC20\n USB: serial: option: add support for Telit LE922 PID 0x1045\n USB: cp210x: Add ID for Parrot NMEA GPS Flight Recorder\n USB: qcserial: add Dell Wireless 5809e Gobi 4G HSPA+ (rev3)\n usb: chipidea: otg: change workqueue ci_otg as freezable\n" }, { "commit": "ad32a1f3c36b046b2e7b0361b88c18e3af61419e", "tree": "18ee25114bc51b98a7b93e370016bbb78fc92ea2", "parents": [ "0834f9cc9f1c5a104e87000378b9cda290c4f406" ], "author": { "name": "Colin Ian King", "email": "colin.king@canonical.com", "time": "Sat Jan 23 19:17:59 2016 +0000" }, "committer": { "name": "Richard Weinberger", "email": "richard@nod.at", "time": "Sat Mar 05 22:21:28 2016 +0100" }, "message": "um: use %lx format specifiers for unsigned longs\n\nstatic analysis from cppcheck detected %x being used for\nunsigned longs:\n\n[arch/x86/um/os-Linux/task_size.c:112]: (warning) %x in format\n string (no. 1) requires \u0027unsigned int\u0027 but the argument type\n is \u0027unsigned long\u0027.\n\nUse %lx instead of %x\n\nSigned-off-by: Colin Ian King \u003ccolin.king@canonical.com\u003e\nSigned-off-by: Richard Weinberger \u003crichard@nod.at\u003e\n" }, { "commit": "0834f9cc9f1c5a104e87000378b9cda290c4f406", "tree": "7ebc8b46ad8beb981358737bcc88ac3915981bc6", "parents": [ "322740efbb7ca26df845161e61cc41484b7e328e" ], "author": { "name": "Richard Weinberger", "email": "richard@nod.at", "time": "Mon Jan 25 23:24:21 2016 +0100" }, "committer": { "name": "Richard Weinberger", "email": "richard@nod.at", "time": "Sat Mar 05 22:17:52 2016 +0100" }, "message": "um: Export pm_power_off\n\n...modules are using this symbol.\nExport it like all other archs to.\n\nSigned-off-by: Richard Weinberger \u003crichard@nod.at\u003e\n" }, { "commit": "322740efbb7ca26df845161e61cc41484b7e328e", "tree": "b8caf74ddac8ab94661a6fed1aad6f65e3194fac", "parents": [ "fc77dbd34c5c99bce46d40a2491937c3bcbd10af" ], "author": { "name": "Richard Weinberger", "email": "richard@nod.at", "time": "Mon Jan 25 23:33:30 2016 +0100" }, "committer": { "name": "Richard Weinberger", "email": "richard@nod.at", "time": "Sat Mar 05 22:16:40 2016 +0100" }, "message": "Revert \"um: Fix get_signal() usage\"\n\nCommit db2f24dc240856fb1d78005307f1523b7b3c121b\nwas plain wrong. I did not realize the we are\nallowed to loop here.\nIn fact we have to loop and must not return to userspace\nbefore all SIGSEGVs have been delivered.\nOther archs do this directly in their entry code, UML\ndoes it here.\n\nReported-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\nSigned-off-by: Richard Weinberger \u003crichard@nod.at\u003e\n" }, { "commit": "e4f6daac20332448529b11f09388f1d55ef2084c", "tree": "176dde53412c2d03a732188850394580159dade8", "parents": [ "fc77dbd34c5c99bce46d40a2491937c3bcbd10af" ], "author": { "name": "Richard Weinberger", "email": "richard@nod.at", "time": "Sun Feb 21 10:53:03 2016 +0100" }, "committer": { "name": "Richard Weinberger", "email": "richard@nod.at", "time": "Sat Mar 05 21:56:23 2016 +0100" }, "message": "ubi: Fix out of bounds write in volume update code\n\nubi_start_leb_change() allocates too few bytes.\nubi_more_leb_change_data() will write up to req-\u003eupd_bytes +\nubi-\u003emin_io_size bytes.\n\nCc: stable@vger.kernel.org\nSigned-off-by: Richard Weinberger \u003crichard@nod.at\u003e\nReviewed-by: Boris Brezillon \u003cboris.brezillon@free-electrons.com\u003e\n" }, { "commit": "67944024c1cdd897e49a09b0d6af3ea38d1388ca", "tree": "b07dfd7bbd04e0d6c2870208802ad7c916f3fb63", "parents": [ "40fea2ed0bcedcab6a7c318f4be2df2bcab4f2eb", "790b415c98de62602810b0eedce26f0f9d6ddd78" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 05 12:35:48 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 05 12:35:48 2016 -0800" }, "message": "Merge tag \u0027sound-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound\n\nPull sound fixes from Takashi Iwai:\n \"It\u0027s our tradition to get a high volume of fixes late at rc7: this\n time, X32 ABI breakage was found and this resulted in a high number\n LOCs. The necessary changes to ALSA core codes were fairly\n straightforward, and more importantly, they are specific to X32, thus\n should be safe to apply.\n\n Other than that, rather a collection of small fixes:\n - Removal of the code that blocks too long at closing the OSS\n sequencer client (which was spotted by syzkaller, unsurprisingly)\n - Fixes races at HD-audio HDMI i915 audio binding\n - a few HDSP/HDPM zero-division fixes\n - Quirks for HD-audio and USB-audio as usual\"\n\n* tag \u0027sound-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:\n ALSA: hda - hdmi defer to register acomp eld notifier\n ALSA: hda - hdmi add wmb barrier for audio component\n ALSA: hda - Fix mic issues on Acer Aspire E1-472\n ALSA: seq: oss: Don\u0027t drain at closing a client\n ALSA: usb-audio: Add a quirk for Plantronics DA45\n ALSA: hdsp: Fix wrong boolean ctl value accesses\n ALSA: hdspm: Fix zero-division\n ALSA: hdspm: Fix wrong boolean ctl value accesses\n ALSA: timer: Fix ioctls for X32 ABI\n ALSA: timer: Fix broken compat timer user status ioctl\n ALSA: rawmidi: Fix ioctls X32 ABI\n ALSA: rawmidi: Use comapt_put_timespec()\n ALSA: pcm: Fix ioctls for X32 ABI\n ALSA: ctl: Fix ioctls for X32 ABI\n" }, { "commit": "40fea2ed0bcedcab6a7c318f4be2df2bcab4f2eb", "tree": "b6f78b8785e7f8f8ab5618e99c87a81aaec9a50b", "parents": [ "ee8f3955c0ef4c7a90a5f707c37c4cb3711ff626", "f16921275cc3c2442d0b95225785a601603b990f" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 05 12:34:29 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 05 12:34:29 2016 -0800" }, "message": "Merge tag \u0027dmaengine-fix-4.5-rc7\u0027 of git://git.infradead.org/users/vkoul/slave-dma\n\nPull dmaengine fix from Vinod Koul:\n \"One minor fix on pxa driver to fix the cyclic dma tranfers\"\n\n* tag \u0027dmaengine-fix-4.5-rc7\u0027 of git://git.infradead.org/users/vkoul/slave-dma:\n dmaengine: pxa_dma: fix cyclic transfers\n" }, { "commit": "ee8f3955c0ef4c7a90a5f707c37c4cb3711ff626", "tree": "ecd1f7ba6771f9b7be9e0fddd37a621bf9988145", "parents": [ "a7c9b603cf2371edacb054abc35597e810c1e5fd", "fbe093ac9f0201939279cdfe8b0fce20ce5ef7a9" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 05 12:32:34 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sat Mar 05 12:32:34 2016 -0800" }, "message": "Merge tag \u0027media/v4.5-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media\n\nPull media fixes from Mauro Carvalho Chehab:\n - some last time changes before we stablize the new entity function\n integer numbers at uAPI\n - probe: fix erroneous return value on i2c/adp1653 driver\n - fix tx 5v detect regression on adv7604 driver\n - fix missing unlock on error in vpfe_prepare_pipeline() on\n davinci_vpfe driver\n\n* tag \u0027media/v4.5-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media:\n [media] media: Sanitise the reserved fields of the G_TOPOLOGY IOCTL arguments\n [media] media.h: postpone connectors entities\n [media] media.h: use hex values for range offsets, move connectors base up.\n [media] adv7604: fix tx 5v detect regression\n [media] media.h: get rid of MEDIA_ENT_F_CONN_TEST\n [media] [for,v4.5] media.h: increase the spacing between function ranges\n [media] media: i2c/adp1653: probe: fix erroneous return value\n [media] media: davinci_vpfe: fix missing unlock on error in vpfe_prepare_pipeline()\n" }, { "commit": "a7c9b603cf2371edacb054abc35597e810c1e5fd", "tree": "ddbe118327a9ce3630bee14d7f69226eee9b6b14", "parents": [ "c12f83c3099c4dd73fe721409d2ce18ca79a753d", "6e2452dff4441e3dc24d415c8b2cda8a3ba52116" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:47:18 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:47:18 2016 -0800" }, "message": "Merge branch \u0027libnvdimm-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm\n\nPull libnvcimm fix from Dan Williams:\n \"One straggling fix for NVDIMM support.\n\n The KVM/QEMU enabling for NVDIMMs has recently reached the point where\n it is able to accept some ACPI _DSM requests from a guest VM. However\n they immediately found that the 4.5-rc kernel is unusable because the\n kernel\u0027s \u0027nfit\u0027 driver fails to load upon seeing a valid \"not\n supported\" response from the virtual BIOS for an address range scrub\n command.\n\n It is not mandatory that a platform implement address range scrubbing,\n so this fix from Vishal properly treats the \u0027not supported\u0027 response\n as \u0027skip scrubbing and continue loading the driver\u0027\"\n\n* \u0027libnvdimm-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm:\n nfit: Continue init even if ARS commands are unimplemented\n" }, { "commit": "c12f83c3099c4dd73fe721409d2ce18ca79a753d", "tree": "3c81b4400d9f281e210eb725d7f5f5779bdf7246", "parents": [ "fab3e94a624f6ac6c4d8b8dc027bb03818a6cf1b", "21b81716c6bff24cda52dc75588455f879ddbfe9" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:41:40 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:41:40 2016 -0800" }, "message": "Merge tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi\n\nPull SCSI fixes from James Bottomley:\n \"Two fairly simple fixes.\n\n One is a regression with ipr firmware loading caused by one of the\n trivial patches in the last merge window which failed to strip the \\n\n from the file name string, so now the firmware loader no longer works\n leading to a lot of unhappy ipr users; fix by stripping the \\n.\n\n The second is a memory leak within SCSI: the BLK_PREP_INVALID state\n was introduced a recent fix but we forgot to account for it correctly\n when freeing state, resulting in memory leakage. Add the correct\n state freeing in scsi_prep_return()\"\n\n* tag \u0027scsi-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:\n ipr: Fix regression when loading firmware\n SCSI: Free resources when we return BLKPREP_INVALID\n" }, { "commit": "fab3e94a624f6ac6c4d8b8dc027bb03818a6cf1b", "tree": "162562a251315c41fabef2d7140f9234ce00abc0", "parents": [ "e5322c54068162846fdbac0f0891cd95f37d4c4e", "dc8b4afc4a04fac8ee55a19b59f2356a25e7e778" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:31:36 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:31:36 2016 -0800" }, "message": "Merge branch \u0027for-4.5-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/libata\n\nPull libata fixes from Tejun Heo:\n \"Assorted fixes for libata drivers.\n\n - Turns out HDIO_GET_32BIT ioctl was subtly broken all along.\n\n - Recent update to ahci external port handling was incorrectly\n marking hotpluggable ports as external making userland handle\n devices connected to those ports incorrectly.\n\n - ahci_xgene needs its own irq handler to work around a hardware\n erratum. libahci updated to allow irq handler override.\n\n - Misc driver specific updates\"\n\n* \u0027for-4.5-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tj/libata:\n ata: ahci: don\u0027t mark HotPlugCapable Ports as external/removable\n ahci: Workaround for ThunderX Errata#22536\n libata: Align ata_device\u0027s id on a cacheline\n Adding Intel Lewisburg device IDs for SATA\n pata-rb532-cf: get rid of the irq_to_gpio() call\n libata: fix HDIO_GET_32BIT ioctl\n ahci_xgene: Implement the workaround to fix the missing of the edge interrupt for the HOST_IRQ_STAT.\n ata: Remove the AHCI_HFLAG_EDGE_IRQ support from libahci.\n libahci: Implement the capability to override the generic ahci interrupt handler.\n" }, { "commit": "e5322c54068162846fdbac0f0891cd95f37d4c4e", "tree": "21d22ea5f1226073e1e106053d16a055be95c387", "parents": [ "bdf9d29799296fd787a9d2b99cb8feff77e2b9db", "4d6af73d9e43f78651a43ee4c5ad221107ac8365" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:17:17 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:17:17 2016 -0800" }, "message": "Merge branch \u0027for-linus2\u0027 of git://git.kernel.dk/linux-block\n\nPull block fixes from Jens Axboe:\n \"Round 2 of this. I cut back to the bare necessities, the patch is\n still larger than it usually would be at this time, due to the number\n of NVMe fixes in there. This pull request contains:\n\n - The 4 core fixes from Ming, that fix both problems with exceeding\n the virtual boundary limit in case of merging, and the gap checking\n for cloned bio\u0027s.\n\n - NVMe fixes from Keith and Christoph:\n\n - Regression on larger user commands, causing problems with\n reading log pages (for instance). This touches both NVMe,\n and the block core since that is now generally utilized also\n for these types of commands.\n\n - Hot removal fixes.\n\n - User exploitable issue with passthrough IO commands, if !length\n is given, causing us to fault on writing to the zero\n page.\n\n - Fix for a hang under error conditions\n\n - And finally, the current series regression for umount with cgroup\n writeback, where the final flush would happen async and hence open\n up window after umount where the device wasn\u0027t consistent. fsck\n right after umount would show this. From Tejun\"\n\n* \u0027for-linus2\u0027 of git://git.kernel.dk/linux-block:\n block: support large requests in blk_rq_map_user_iov\n block: fix blk_rq_get_max_sectors for driver private requests\n nvme: fix max_segments integer truncation\n nvme: set queue limits for the admin queue\n writeback: flush inode cgroup wb switches instead of pinning super_block\n NVMe: Fix 0-length integrity payload\n NVMe: Don\u0027t allow unsupported flags\n NVMe: Move error handling to failed reset handler\n NVMe: Simplify device reset failure\n NVMe: Fix namespace removal deadlock\n NVMe: Use IDA for namespace disk naming\n NVMe: Don\u0027t unmap controller registers on reset\n block: merge: get the 1st and last bvec via helpers\n block: get the 1st and last bvec via helpers\n block: check virt boundary in bio_will_gap()\n block: bio: introduce helpers to get the 1st and last bvec\n" }, { "commit": "bdf9d29799296fd787a9d2b99cb8feff77e2b9db", "tree": "376675a648e7f50beb6a36b0f9b9aad37bb73cd0", "parents": [ "638c201e3fffd2699e58305030d58730355ece12", "11d8d645343efba0c975aefe7c2cf3b33c836c75" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:06:49 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 18:06:49 2016 -0800" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma\n\nPull rdma fixes from Doug Ledford:\n \"Additional 4.5-rc6 fixes.\n\n I have four patches today. I had previously thought I had submitted\n two of them last week, but they were accidentally skipped :-(.\n\n - One fix to an error path in the core\n - One fix for RoCE in the core\n - Two related fixes for the core/mlx5\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma:\n IB/core: Use GRH when the path hop-limit \u003e 0\n IB/{core, mlx5}: Fix input len in vendor part of create_qp/srq\n IB/mlx5: Avoid using user-index for SRQs\n IB/core: Fix missed clean call in registration path\n" }, { "commit": "638c201e3fffd2699e58305030d58730355ece12", "tree": "5d44c958d1a0e341dbdfbfd1d5301f2a4ce76023", "parents": [ "b80e8e2811ecb2ace7ff851d887fe8b6db4425bf", "26bae5e04c3570728787f809b13546a8169fdf95" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:56:48 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:56:48 2016 -0800" }, "message": "Merge branch \u0027drm-fixes\u0027 of git://people.freedesktop.org/~airlied/linux\n\nPull drm fixes from Dave Airlie:\n \"This contains one i915 patch twice, as I merged it locally for\n testing, and then pulled some stuff in on top, and then Jani sent to\n me, I didn\u0027t think it was worth redoing all the merges of what I had\n tested.\n\n Summary:\n\n - amdgpu/radeon fixes for some more power management and VM races.\n\n - Two i915 fixes, one for the a recent regression, one another power\n management fix for skylake.\n\n - Two tegra dma mask fixes for a regression.\n\n - One ast fix for a typo I made transcribing the userspace driver,\n that I\u0027d like to get into stable so I don\u0027t forget about it\"\n\n* \u0027drm-fixes\u0027 of git://people.freedesktop.org/~airlied/linux:\n gpu: host1x: Set DMA ops on device creation\n gpu: host1x: Set DMA mask\n drm/amdgpu: return from atombios_dp_get_dpcd only when error\n drm/amdgpu/cz: remove commented out call to enable vce pg\n drm/amdgpu/powerplay/cz: enable/disable vce dpm independent of vce pg\n drm/amdgpu/cz: enable/disable vce dpm even if vce pg is disabled\n drm/amdgpu/gfx8: specify which engine to wait before vm flush\n drm/amdgpu: apply gfx_v8 fixes to gfx_v7 as well\n drm/amd/powerplay: send event to notify powerplay all modules are initialized.\n drm/amd/powerplay: export AMD_PP_EVENT_COMPLETE_INIT task to amdgpu.\n drm/radeon/pm: update current crtc info after setting the powerstate\n drm/amdgpu/pm: update current crtc info after setting the powerstate\n drm/i915: Balance assert_rpm_wakelock_held() for !IS_ENABLED(CONFIG_PM)\n drm/i915/skl: Fix power domain suspend sequence\n drm/ast: Fix incorrect register check for DRAM width\n drm/i915: Balance assert_rpm_wakelock_held() for !IS_ENABLED(CONFIG_PM)\n" }, { "commit": "b80e8e2811ecb2ace7ff851d887fe8b6db4425bf", "tree": "057459dc4eaa157e1894e02d14b08a1b3e2a0296", "parents": [ "ed385c7a1752e6a87622e5a2a80808cfb8917305", "bfc6b97d528def357adcbdadf71b21c8e5f7dbc4" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:51:16 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:51:16 2016 -0800" }, "message": "Merge tag \u0027pm+acpi-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm\n\nPull power management and ACPI fixes from Rafael Wysocki:\n \"Two build fixes for cpufreq drivers (including one for breakage\n introduced recently) and a fix for a graph tracer crash when used over\n suspend-to-RAM on x86.\n\n Specifics:\n\n - Prevent the graph tracer from crashing when used over suspend-to-\n RAM on x86 by pausing it before invoking do_suspend_lowlevel() and\n un-pausing it when that function has returned (Todd Brandt).\n\n - Fix build issues in the qoriq and mediatek cpufreq drivers related\n to broken dependencies on THERMAL (Arnd Bergmann)\"\n\n* tag \u0027pm+acpi-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm:\n PM / sleep / x86: Fix crash on graph trace through x86 suspend\n cpufreq: mediatek: allow building as a module\n cpufreq: qoriq: allow building as module with THERMAL\u003dm\n" }, { "commit": "ed385c7a1752e6a87622e5a2a80808cfb8917305", "tree": "836d8197351ff0cc1187536d466bd6bbea77bd72", "parents": [ "c51797d25d84833804e9e6950c8c1abbe6c73808", "dfd55ad85e4a7fbaa82df12467515ac3c81e8a3e" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:43:40 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:43:40 2016 -0800" }, "message": "Merge tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux\n\nPull arm64 fix from Will Deacon:\n \"Arm64 fix for -rc7. Without it, our struct page array can overflow\n the vmemmap region on systems with a large PHYS_OFFSET.\n\n Nothing else on the radar at the moment, so hopefully that\u0027s it for\n 4.5 from us.\n\n Summary: Ensure struct page array fits within vmemmap area\"\n\n* tag \u0027arm64-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:\n arm64: vmemmap: use virtual projection of linear region\n" }, { "commit": "c51797d25d84833804e9e6950c8c1abbe6c73808", "tree": "4b43585feca6e7d76cd67f51799de199ccb43d77", "parents": [ "2cdcb2b5b5d6d7c1bdefbc1a63187d47fd809408", "44248affb5d82d32cdb58a92a94ce191d4ddee60" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:36:46 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:36:46 2016 -0800" }, "message": "Merge tag \u0027for-linus-20160304\u0027 of git://git.infradead.org/linux-mtd\n\nPull jffs2 fixes from David Woodhouse:\n \"This contains two important JFFS2 fixes marked for stable:\n\n - a lock ordering problem between the page lock and the internal\n f-\u003esem mutex, which was causing occasional deadlocks in garbage\n collection\n\n - a scan failure causing moved directories to sometimes end up\n appearing to have hard links.\n\n There are also a couple of trivial MAINTAINERS file updates\"\n\n* tag \u0027for-linus-20160304\u0027 of git://git.infradead.org/linux-mtd:\n MAINTAINERS: add maintainer entry for FREESCALE GPMI NAND driver\n Fix directory hardlinks from deleted directories\n jffs2: Fix page lock / f-\u003esem deadlock\n Revert \"jffs2: Fix lock acquisition order bug in jffs2_write_begin\"\n MAINTAINERS: update Han\u0027s email\n" }, { "commit": "2cdcb2b5b5d6d7c1bdefbc1a63187d47fd809408", "tree": "21d2465f170867584a2a0e5a8a9bd5227424ec82", "parents": [ "78baab7aa843ef01644195d341bee118cbe4bf19", "909c3a22da3b8d2cfd3505ca5658f0176859d400" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:31:32 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 17:31:32 2016 -0800" }, "message": "Merge branch \u0027for-linus-4.5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs\n\nPull btrfs fix from Chris Mason:\n \"Filipe nailed down a problem where tree log replay would do some work\n that orphan code wasn\u0027t expecting to be done yet, leading to BUG_ON\"\n\n* \u0027for-linus-4.5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:\n Btrfs: fix loading of orphan roots leading to BUG_ON\n" }, { "commit": "78baab7aa843ef01644195d341bee118cbe4bf19", "tree": "73502edec6885609de4bd359cc0fc1b33af7a333", "parents": [ "e3c2ef41f88e50c8557270868600d3132028af3b", "e57cbaf0eb006eaa207395f3bfd7ce52c1b5539c" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 16:57:04 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 04 16:57:04 2016 -0800" }, "message": "Merge tag \u0027trace-fixes-v4.5-rc6\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace\n\nPull tracing fix from Steven Rostedt:\n \"A feature was added in 4.3 that allowed users to filter trace points\n on a tasks \"comm\" field. But this prevented filtering on a comm field\n that is within a trace event (like sched_migrate_task).\n\n When trying to filter on when a program migrated, this change\n prevented the filtering of the sched_migrate_task.\n\n To fix this, the event fields are examined first, and then the extra\n fields like \"comm\" and \"cpu\" are examined. Also, instead of testing\n to assign the comm filter function based on the field\u0027s name, the\n generic comm field is given a new filter type (FILTER_COMM). When\n this field is used to filter the type is checked. The same is done\n for the cpu filter field.\n\n Two new special filter types are added: \"COMM\" and \"CPU\". This allows\n users to still filter the tasks comm for events that have \"comm\" as\n one of their fields, in cases that users would like to filter\n sched_migrate_task on the comm of the task that called the event, and\n not the comm of the task that is being migrated\"\n\n* tag \u0027trace-fixes-v4.5-rc6\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:\n tracing: Do not have \u0027comm\u0027 filter override event \u0027comm\u0027 field\n" }, { "commit": "6e2452dff4441e3dc24d415c8b2cda8a3ba52116", "tree": "d4f99e6ee0506b19e93220f8b5031ae2e9ea3286", "parents": [ "fc77dbd34c5c99bce46d40a2491937c3bcbd10af" ], "author": { "name": "Vishal Verma", "email": "vishal.l.verma@intel.com", "time": "Thu Mar 03 15:39:41 2016 -0700" }, "committer": { "name": "Dan Williams", "email": "dan.j.williams@intel.com", "time": "Fri Mar 04 16:46:13 2016 -0800" }, "message": "nfit: Continue init even if ARS commands are unimplemented\n\nIf firmware doesn\u0027t implement any of the ARS commands, take that to\nmean that ARS is unsupported, and continue to initialize regions without\nbad block lists. We cannot make the assumption that ARS commands will be\nunconditionally supported on all NVDIMMs.\n\nReported-by: Haozhong Zhang \u003chaozhong.zhang@intel.com\u003e\nSigned-off-by: Vishal Verma \u003cvishal.l.verma@intel.com\u003e\nAcked-by: Xiao Guangrong \u003cguangrong.xiao@linux.intel.com\u003e\nTested-by: Haozhong Zhang \u003chaozhong.zhang@intel.com\u003e\nSigned-off-by: Dan Williams \u003cdan.j.williams@intel.com\u003e\n" }, { "commit": "f474c8c857d996f34c39f66bbed23faaa739fad6", "tree": "0a0ae7bb3a9a7c405d6bba7cca7b75dabe5ea1fc", "parents": [ "c7edd7f99cdaec65e9303b4c9edc19bb595564ed" ], "author": { "name": "Mika Penttilä", "email": "mika.penttila@nextfour.com", "time": "Mon Feb 22 17:56:52 2016 +0100" }, "committer": { "name": "Russell King", "email": "rmk+kernel@arm.linux.org.uk", "time": "Fri Mar 04 23:32:45 2016 +0000" }, "message": "ARM: 8544/1: set_memory_xx fixes\n\nAllow zero size updates. This makes set_memory_xx() consistent with x86, s390 and arm64 and makes apply_to_page_range() not to BUG() when loading modules.\n\nSigned-off-by: Mika Penttilä mika.penttila@nextfour.com\nSigned-off-by: Russell King \u003crmk+kernel@arm.linux.org.uk\u003e\n" }, { "commit": "26bae5e04c3570728787f809b13546a8169fdf95", "tree": "ddb961dc8fec60bbe73ae97c4a37df8f244df1aa", "parents": [ "0dff9738ec6dddf701821a6aa477ced809f17ef0", "c95469aa5a188384ccf8ac520ece931c66caf8aa" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sat Mar 05 07:53:25 2016 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sat Mar 05 07:53:25 2016 +1000" }, "message": "Merge tag \u0027drm/tegra/for-4.5-rc7\u0027 of git://anongit.freedesktop.org/tegra/linux into drm-fixes\n\ndrm/tegra: Fixes for v4.5-rc7\n\nTwo small fixes that restore PRIME support.\n\n* tag \u0027drm/tegra/for-4.5-rc7\u0027 of git://anongit.freedesktop.org/tegra/linux:\n gpu: host1x: Set DMA ops on device creation\n gpu: host1x: Set DMA mask\n" }, { "commit": "e723e3f7f9591b79e8c56b3d7c5a204a9c571b55", "tree": "24f595ea619d1febe04a97e97ea8536bdb11ed92", "parents": [ "887349f69f37e71e2a8bfbd743831625a0b2ff51" ], "author": { "name": "Maciej W. Rozycki", "email": "macro@imgtec.com", "time": "Fri Mar 04 01:42:49 2016 +0000" }, "committer": { "name": "Ralf Baechle", "email": "ralf@linux-mips.org", "time": "Fri Mar 04 22:52:32 2016 +0100" }, "message": "MIPS: traps: Fix SIGFPE information leak from `do_ov\u0027 and `do_trap_or_bp\u0027\n\nAvoid sending a partially initialised `siginfo_t\u0027 structure along SIGFPE\nsignals issued from `do_ov\u0027 and `do_trap_or_bp\u0027, leading to information\nleaking from the kernel stack.\n\nSigned-off-by: Maciej W. Rozycki \u003cmacro@imgtec.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Ralf Baechle \u003cralf@linux-mips.org\u003e\n" }, { "commit": "bfc6b97d528def357adcbdadf71b21c8e5f7dbc4", "tree": "3fe087169320a608b62decc7bf28dbb003b499ec", "parents": [ "fc77dbd34c5c99bce46d40a2491937c3bcbd10af", "3c2002aec3ee3a9fc86ff8f8618f1253c94ec919", "92f9e179a702a6adbc11e2fedc76ecd6ffc9e3f7" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Fri Mar 04 22:41:05 2016 +0100" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Fri Mar 04 22:41:05 2016 +0100" }, "message": "Merge branches \u0027pm-cpufreq-fixes\u0027 and \u0027pm-sleep-fixes\u0027\n\n* pm-cpufreq-fixes:\n cpufreq: mediatek: allow building as a module\n cpufreq: qoriq: allow building as module with THERMAL\u003dm\n\n* pm-sleep-fixes:\n PM / sleep / x86: Fix crash on graph trace through x86 suspend\n" }, { "commit": "5ea5c5e0a7f70b256417d3b6e36bd9851504babd", "tree": "c71678dee6860e0bce59c737a3ea52f7e310e09f", "parents": [ "fc77dbd34c5c99bce46d40a2491937c3bcbd10af" ], "author": { "name": "Yan, Zheng", "email": "zyan@redhat.com", "time": "Sun Feb 14 18:06:41 2016 +0800" }, "committer": { "name": "Ilya Dryomov", "email": "idryomov@gmail.com", "time": "Fri Mar 04 21:00:37 2016 +0100" }, "message": "ceph: initial CEPH_FEATURE_FS_FILE_LAYOUT_V2 support\n\nAdd support for the format change of MClientReply/MclientCaps.\nAlso add code that denies access to inodes with pool_ns layouts.\n\nSigned-off-by: Yan, Zheng \u003czyan@redhat.com\u003e\nReviewed-by: Sage Weil \u003csage@redhat.com\u003e\n" }, { "commit": "c95469aa5a188384ccf8ac520ece931c66caf8aa", "tree": "ff5b147057b1b5f9fb981ae411c173865e0a40d2", "parents": [ "097452e61366a939a4772332181cea7cdcc74760" ], "author": { "name": "Alexandre Courbot", "email": "acourbot@nvidia.com", "time": "Fri Feb 26 18:06:53 2016 +0900" }, "committer": { "name": "Thierry Reding", "email": "treding@nvidia.com", "time": "Fri Mar 04 16:24:57 2016 +0100" }, "message": "gpu: host1x: Set DMA ops on device creation\n\nCurrently host1x-instanciated devices have their dma_ops left to NULL,\nwhich makes any DMA operation (like buffer import) on ARM64 fallback\nto the dummy_dma_ops and fail with an error.\n\nThis patch calls of_dma_configure() with the host1x node when creating\nsuch a device, so the proper DMA operations are set.\n\nSuggested-by: Thierry Reding \u003cthierry.reding@gmail.com\u003e\nSigned-off-by: Alexandre Courbot \u003cacourbot@nvidia.com\u003e\nSigned-off-by: Thierry Reding \u003ctreding@nvidia.com\u003e\n" }, { "commit": "097452e61366a939a4772332181cea7cdcc74760", "tree": "ba2f85ae6a48c6930b51e8cef4c2a691bb4f788a", "parents": [ "92e963f50fc74041b5e9e744c330dca48e04f08d" ], "author": { "name": "Alexandre Courbot", "email": "acourbot@nvidia.com", "time": "Fri Feb 26 18:06:52 2016 +0900" }, "committer": { "name": "Thierry Reding", "email": "treding@nvidia.com", "time": "Fri Mar 04 16:24:56 2016 +0100" }, "message": "gpu: host1x: Set DMA mask\n\nThe default DMA mask covers a 32 bits address range, but host1x devices\ncan address a larger range on TK1 and TX1. Set the DMA mask to the range\naddressable when we use the IOMMU to prevent the use of bounce buffers.\n\nSigned-off-by: Alexandre Courbot \u003cacourbot@nvidia.com\u003e\nSigned-off-by: Thierry Reding \u003ctreding@nvidia.com\u003e\n" }, { "commit": "e57cbaf0eb006eaa207395f3bfd7ce52c1b5539c", "tree": "a3ec2acf045e5f9ec191f391a501fae096f4ca9a", "parents": [ "d045437a169f899dfb0f6f7ede24cc042543ced9" ], "author": { "name": "Steven Rostedt (Red Hat)", "email": "rostedt@goodmis.org", "time": "Thu Mar 03 17:18:20 2016 -0500" }, "committer": { "name": "Steven Rostedt", "email": "rostedt@goodmis.org", "time": "Fri Mar 04 09:57:10 2016 -0500" }, "message": "tracing: Do not have \u0027comm\u0027 filter override event \u0027comm\u0027 field\n\nCommit 9f61668073a8d \"tracing: Allow triggers to filter for CPU ids and\nprocess names\" added a \u0027comm\u0027 filter that will filter events based on the\ncurrent tasks struct \u0027comm\u0027. But this now hides the ability to filter events\nthat have a \u0027comm\u0027 field too. For example, sched_migrate_task trace event.\nThat has a \u0027comm\u0027 field of the task to be migrated.\n\n echo \u0027comm \u003d\u003d \"bash\"\u0027 \u003e events/sched_migrate_task/filter\n\nwill now filter all sched_migrate_task events for tasks named \"bash\" that\nmigrates other tasks (in interrupt context), instead of seeing when \"bash\"\nitself gets migrated.\n\nThis fix requires a couple of changes.\n\n1) Change the look up order for filter predicates to look at the events\n fields before looking at the generic filters.\n\n2) Instead of basing the filter function off of the \"comm\" name, have the\n generic \"comm\" filter have its own filter_type (FILTER_COMM). Test\n against the type instead of the name to assign the filter function.\n\n3) Add a new \"COMM\" filter that works just like \"comm\" but will filter based\n on the current task, even if the trace event contains a \"comm\" field.\n\nDo the same for \"cpu\" field, adding a FILTER_CPU and a filter \"CPU\".\n\nCc: stable@vger.kernel.org # v4.3+\nFixes: 9f61668073a8d \"tracing: Allow triggers to filter for CPU ids and process names\"\nReported-by: Matt Fleming \u003cmatt@codeblueprint.co.uk\u003e\nSigned-off-by: Steven Rostedt \u003crostedt@goodmis.org\u003e\n" }, { "commit": "790b415c98de62602810b0eedce26f0f9d6ddd78", "tree": "31eae524fe8e36df55b07c844a4c2e21b0d312b3", "parents": [ "ec75a940b1037e877efd9a5a9e94eab1e464f73b" ], "author": { "name": "Libin Yang", "email": "libin.yang@linux.intel.com", "time": "Fri Mar 04 14:33:43 2016 +0800" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Fri Mar 04 08:40:24 2016 +0100" }, "message": "ALSA: hda - hdmi defer to register acomp eld notifier\n\nDefer to register acomp eld notifier until hdmi audio driver\nis fully ready.\n\nAfter registering eld notifier, gfx driver can use this\ncallback function to notify audio driver the monitor\nconnection event. However this action may happen when\naudio driver is adding the pins or doing other initialization.\nThis is not always safe, however. For example, using\nper_pin-\u003elock before the lock is initialized.\n\nLet\u0027s register the eld notifier after the initialization is done.\n\nSigned-off-by: Libin Yang \u003clibin.yang@linux.intel.com\u003e\nSigned-off-by: Takashi Iwai \u003ctiwai@suse.de\u003e\n" }, { "commit": "ec75a940b1037e877efd9a5a9e94eab1e464f73b", "tree": "10e876cc4992497a328580ab7446579e1ee01c21", "parents": [ "02322ac9dee9aff8d8862e8d6660ebe102f492ea" ], "author": { "name": "Libin Yang", "email": "libin.yang@linux.intel.com", "time": "Fri Mar 04 14:33:06 2016 +0800" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Fri Mar 04 08:37:59 2016 +0100" }, "message": "ALSA: hda - hdmi add wmb barrier for audio component\n\nTo make sure audio_ptr is set before intel_audio_codec_enable()\nor intel_audio_codec_disable() calling pin_eld_notify(),\nthis patch adds wmb barrier to prevent optimizing.\n\nSigned-off-by: Libin Yang \u003clibin.yang@linux.intel.com\u003e\nSigned-off-by: Takashi Iwai \u003ctiwai@suse.de\u003e\n" }, { "commit": "37c5e942bb2eedd98c1cd1fa1f94d79f6b830c38", "tree": "2d846e3812b414f352d91779c311e4b08f1cd069", "parents": [ "fb822e6076d972691c5dd33431c10f82dda94ae9" ], "author": { "name": "Scott Wood", "email": "oss@buserror.net", "time": "Wed Mar 02 22:51:04 2016 -0600" }, "committer": { "name": "Scott Wood", "email": "oss@buserror.net", "time": "Thu Mar 03 23:43:05 2016 -0600" }, "message": "powerpc/fsl-book3e: Avoid lbarx on e5500\n\nlbarx/stbcx. are implemented on e6500, but not on e5500.\nLikewise, SMT is on e6500, but not on e5500.\n\nSo, avoid executing an unimplemented instruction by only locking\nwhen needed (i.e. in the presence of SMT).\n\nSigned-off-by: Scott Wood \u003coss@buserror.net\u003e\n" }, { "commit": "0dff9738ec6dddf701821a6aa477ced809f17ef0", "tree": "3853bfbcb6790bfe7866abda9238d62136e86afa", "parents": [ "f0511e66114a6414cfca92d7b94118913a0c11ff", "eda908967feecf7575a8ab74f1695b0445cf324e" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 04 13:51:53 2016 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Fri Mar 04 13:51:53 2016 +1000" }, "message": "Merge tag \u0027drm-intel-fixes-2016-03-03\u0027 of git://anongit.freedesktop.org/drm-intel into drm-fixes\n\nSmall conflict as I had the balance in my tree already for testing.\n\n* tag \u0027drm-intel-fixes-2016-03-03\u0027 of git://anongit.freedesktop.org/drm-intel:\n drm/i915: Balance assert_rpm_wakelock_held() for !IS_ENABLED(CONFIG_PM)\n drm/i915/skl: Fix power domain suspend sequence\n" }, { "commit": "909c3a22da3b8d2cfd3505ca5658f0176859d400", "tree": "451d03827a4dc4e006f1f4c0022bfdd3c042ca86", "parents": [ "413eddc6ca748dc52057f142d707b9378af692fc" ], "author": { "name": "Filipe Manana", "email": "fdmanana@suse.com", "time": "Wed Mar 02 15:49:38 2016 +0000" }, "committer": { "name": "Chris Mason", "email": "clm@fb.com", "time": "Thu Mar 03 15:28:59 2016 -0800" }, "message": "Btrfs: fix loading of orphan roots leading to BUG_ON\n\nWhen looking for orphan roots during mount we can end up hitting a\nBUG_ON() (at root-item.c:btrfs_find_orphan_roots()) if a log tree is\nreplayed and qgroups are enabled. This is because after a log tree is\nreplayed, a transaction commit is made, which triggers qgroup extent\naccounting which in turn does backref walking which ends up reading and\ninserting all roots in the radix tree fs_info-\u003efs_root_radix, including\norphan roots (deleted snapshots). So after the log tree is replayed, when\nfinding orphan roots we hit the BUG_ON with the following trace:\n\n[118209.182438] ------------[ cut here ]------------\n[118209.183279] kernel BUG at fs/btrfs/root-tree.c:314!\n[118209.184074] invalid opcode: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC\n[118209.185123] Modules linked in: btrfs dm_flakey dm_mod crc32c_generic ppdev xor raid6_pq evdev sg parport_pc parport acpi_cpufreq tpm_tis tpm psmouse\nprocessor i2c_piix4 serio_raw pcspkr i2c_core button loop autofs4 ext4 crc16 mbcache jbd2 sd_mod sr_mod cdrom ata_generic virtio_scsi ata_piix libata\nvirtio_pci virtio_ring virtio scsi_mod e1000 floppy [last unloaded: btrfs]\n[118209.186318] CPU: 14 PID: 28428 Comm: mount Tainted: G W 4.5.0-rc5-btrfs-next-24+ #1\n[118209.186318] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS by qemu-project.org 04/01/2014\n[118209.186318] task: ffff8801ec131040 ti: ffff8800af34c000 task.ti: ffff8800af34c000\n[118209.186318] RIP: 0010:[\u003cffffffffa04237d7\u003e] [\u003cffffffffa04237d7\u003e] btrfs_find_orphan_roots+0x1fc/0x244 [btrfs]\n[118209.186318] RSP: 0018:ffff8800af34faa8 EFLAGS: 00010246\n[118209.186318] RAX: 00000000ffffffef RBX: 00000000ffffffef RCX: 0000000000000001\n[118209.186318] RDX: 0000000080000000 RSI: 0000000000000001 RDI: 00000000ffffffff\n[118209.186318] RBP: ffff8800af34fb08 R08: 0000000000000001 R09: 0000000000000000\n[118209.186318] R10: ffff8800af34f9f0 R11: 6db6db6db6db6db7 R12: ffff880171b97000\n[118209.186318] R13: ffff8801ca9d65e0 R14: ffff8800afa2e000 R15: 0000160000000000\n[118209.186318] FS: 00007f5bcb914840(0000) GS:ffff88023edc0000(0000) knlGS:0000000000000000\n[118209.186318] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b\n[118209.186318] CR2: 00007f5bcaceb5d9 CR3: 00000000b49b5000 CR4: 00000000000006e0\n[118209.186318] Stack:\n[118209.186318] fffffbffffffffff 010230ffffffffff 0101000000000000 ff84000000000000\n[118209.186318] fbffffffffffffff 30ffffffffffffff 0000000000000101 ffff880082348000\n[118209.186318] 0000000000000000 ffff8800afa2e000 ffff8800afa2e000 0000000000000000\n[118209.186318] Call Trace:\n[118209.186318] [\u003cffffffffa042e2db\u003e] open_ctree+0x1e37/0x21b9 [btrfs]\n[118209.186318] [\u003cffffffffa040a753\u003e] btrfs_mount+0x97e/0xaed [btrfs]\n[118209.186318] [\u003cffffffff8108e1c0\u003e] ? trace_hardirqs_on+0xd/0xf\n[118209.186318] [\u003cffffffff8117b87e\u003e] mount_fs+0x67/0x131\n[118209.186318] [\u003cffffffff81192d2b\u003e] vfs_kern_mount+0x6c/0xde\n[118209.186318] [\u003cffffffffa0409f81\u003e] btrfs_mount+0x1ac/0xaed [btrfs]\n[118209.186318] [\u003cffffffff8108e1c0\u003e] ? trace_hardirqs_on+0xd/0xf\n[118209.186318] [\u003cffffffff8108c26b\u003e] ? lockdep_init_map+0xb9/0x1b3\n[118209.186318] [\u003cffffffff8117b87e\u003e] mount_fs+0x67/0x131\n[118209.186318] [\u003cffffffff81192d2b\u003e] vfs_kern_mount+0x6c/0xde\n[118209.186318] [\u003cffffffff81195637\u003e] do_mount+0x8a6/0x9e8\n[118209.186318] [\u003cffffffff8119598d\u003e] SyS_mount+0x77/0x9f\n[118209.186318] [\u003cffffffff81493017\u003e] entry_SYSCALL_64_fastpath+0x12/0x6b\n[118209.186318] Code: 64 00 00 85 c0 89 c3 75 24 f0 41 80 4c 24 20 20 49 8b bc 24 f0 01 00 00 4c 89 e6 e8 e8 65 00 00 85 c0 89 c3 74 11 83 f8 ef 75 02 \u003c0f\u003e 0b\n4c 89 e7 e8 da 72 00 00 eb 1c 41 83 bc 24 00 01 00 00 00\n[118209.186318] RIP [\u003cffffffffa04237d7\u003e] btrfs_find_orphan_roots+0x1fc/0x244 [btrfs]\n[118209.186318] RSP \u003cffff8800af34faa8\u003e\n[118209.230735] ---[ end trace 83938f987d85d477 ]---\n\nSo fix this by not treating the error -EEXIST, returned when attempting\nto insert a root already inserted by the backref walking code, as an error.\n\nThe following test case for xfstests reproduces the bug:\n\n seq\u003d`basename $0`\n seqres\u003d$RESULT_DIR/$seq\n echo \"QA output created by $seq\"\n tmp\u003d/tmp/$$\n status\u003d1\t# failure is the default!\n trap \"_cleanup; exit \\$status\" 0 1 2 3 15\n\n _cleanup()\n {\n _cleanup_flakey\n cd /\n rm -f $tmp.*\n }\n\n # get standard environment, filters and checks\n . ./common/rc\n . ./common/filter\n . ./common/dmflakey\n\n # real QA test starts here\n _supported_fs btrfs\n _supported_os Linux\n _require_scratch\n _require_dm_target flakey\n _require_metadata_journaling $SCRATCH_DEV\n\n rm -f $seqres.full\n\n _scratch_mkfs \u003e\u003e$seqres.full 2\u003e\u00261\n _init_flakey\n _mount_flakey\n\n _run_btrfs_util_prog quota enable $SCRATCH_MNT\n\n # Create 2 directories with one file in one of them.\n # We use these just to trigger a transaction commit later, moving the file from\n # directory a to directory b and doing an fsync against directory a.\n mkdir $SCRATCH_MNT/a\n mkdir $SCRATCH_MNT/b\n touch $SCRATCH_MNT/a/f\n sync\n\n # Create our test file with 2 4K extents.\n $XFS_IO_PROG -f -s -c \"pwrite -S 0xaa 0 8K\" $SCRATCH_MNT/foobar | _filter_xfs_io\n\n # Create a snapshot and delete it. This doesn\u0027t really delete the snapshot\n # immediately, just makes it inaccessible and invisible to user space, the\n # snapshot is deleted later by a dedicated kernel thread (cleaner kthread)\n # which is woke up at the next transaction commit.\n # A root orphan item is inserted into the tree of tree roots, so that if a\n # power failure happens before the dedicated kernel thread does the snapshot\n # deletion, the next time the filesystem is mounted it resumes the snapshot\n # deletion.\n _run_btrfs_util_prog subvolume snapshot $SCRATCH_MNT $SCRATCH_MNT/snap\n _run_btrfs_util_prog subvolume delete $SCRATCH_MNT/snap\n\n # Now overwrite half of the extents we wrote before. Because we made a snapshpot\n # before, which isn\u0027t really deleted yet (since no transaction commit happened\n # after we did the snapshot delete request), the non overwritten extents get\n # referenced twice, once by the default subvolume and once by the snapshot.\n $XFS_IO_PROG -c \"pwrite -S 0xbb 4K 8K\" $SCRATCH_MNT/foobar | _filter_xfs_io\n\n # Now move file f from directory a to directory b and fsync directory a.\n # The fsync on the directory a triggers a transaction commit (because a file\n # was moved from it to another directory) and the file fsync leaves a log tree\n # with file extent items to replay.\n mv $SCRATCH_MNT/a/f $SCRATCH_MNT/a/b\n $XFS_IO_PROG -c \"fsync\" $SCRATCH_MNT/a\n $XFS_IO_PROG -c \"fsync\" $SCRATCH_MNT/foobar\n\n echo \"File digest before power failure:\"\n md5sum $SCRATCH_MNT/foobar | _filter_scratch\n\n # Now simulate a power failure and mount the filesystem to replay the log tree.\n # After the log tree was replayed, we used to hit a BUG_ON() when processing\n # the root orphan item for the deleted snapshot. This is because when processing\n # an orphan root the code expected to be the first code inserting the root into\n # the fs_info-\u003efs_root_radix radix tree, while in reallity it was the second\n # caller attempting to do it - the first caller was the transaction commit that\n # took place after replaying the log tree, when updating the qgroup counters.\n _flakey_drop_and_remount\n\n echo \"File digest before after failure:\"\n # Must match what he got before the power failure.\n md5sum $SCRATCH_MNT/foobar | _filter_scratch\n\n _unmount_flakey\n status\u003d0\n exit\n\nFixes: 2d9e97761087 (\"Btrfs: use btrfs_get_fs_root in resolve_indirect_ref\")\nCc: stable@vger.kernel.org # 4.4+\nSigned-off-by: Filipe Manana \u003cfdmanana@suse.com\u003e\nReviewed-by: Qu Wenruo \u003cquwenruo@cn.fujitsu.com\u003e\nSigned-off-by: Chris Mason \u003cclm@fb.com\u003e\n" }, { "commit": "4d6af73d9e43f78651a43ee4c5ad221107ac8365", "tree": "4f267be22aec6c45e24894be81cc92c3bc1715f5", "parents": [ "f21018427cb007a0894c36ad702990ab639cbbb4" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Mar 02 18:07:14 2016 +0100" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:45:02 2016 -0700" }, "message": "block: support large requests in blk_rq_map_user_iov\n\nThis patch adds support for larger requests in blk_rq_map_user_iov by\nallowing it to build multiple bios for a request. This functionality\nused to exist for the non-vectored blk_rq_map_user in the past, and\nthis patch reuses the existing functionality for it on the unmap side,\nwhich stuck around. Thanks to the iov_iter API supporting multiple\nbios is fairly trivial, as we can just iterate the iov until we\u0027ve\nconsumed the whole iov_iter.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReported-by: Jeff Lien \u003cJeff.Lien@hgst.com\u003e\nTested-by: Jeff Lien \u003cJeff.Lien@hgst.com\u003e\nReviewed-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "f21018427cb007a0894c36ad702990ab639cbbb4", "tree": "88eff3b50155c51a210c61b0805f5a9c0daa9a14", "parents": [ "45686b6198bd824f083ff5293f191d78db9d708a" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Thu Mar 03 14:43:45 2016 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:43:45 2016 -0700" }, "message": "block: fix blk_rq_get_max_sectors for driver private requests\n\nDriver private request types should not get the artifical cap for the\nFS requests. This is important to use the full device capabilities\nfor internal command or NVMe pass through commands.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReported-by: Jeff Lien \u003cJeff.Lien@hgst.com\u003e\nTested-by: Jeff Lien \u003cJeff.Lien@hgst.com\u003e\nReviewed-by: Keith Busch \u003ckeith.busch@intel.com\u003e\n\nUpdated by me to use an explicit check for the one command type that\ndoes support extended checking, instead of relying on the ordering\nof the enum command values - as suggested by Keith.\n\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "45686b6198bd824f083ff5293f191d78db9d708a", "tree": "5c752a162d53af164942242d217d4cf54a98afd8", "parents": [ "da35825d9a091a7a1d5824c8468168e2658333ff" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Mar 02 18:07:12 2016 +0100" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:43:10 2016 -0700" }, "message": "nvme: fix max_segments integer truncation\n\nThe block layer uses an unsigned short for max_segments. The way we\ncalculate the value for NVMe tends to generate very large 32-bit values,\nwhich after integer truncation may lead to a zero value instead of\nthe desired outcome.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReported-by: Jeff Lien \u003cJeff.Lien@hgst.com\u003e\nTested-by: Jeff Lien \u003cJeff.Lien@hgst.com\u003e\nReviewed-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "da35825d9a091a7a1d5824c8468168e2658333ff", "tree": "a79f4c7eaa2d392a6f09f93f691eeea7ba1e479e", "parents": [ "a1a0e23e49037c23ea84bc8cc146a03584d13577" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed Mar 02 18:07:11 2016 +0100" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:50 2016 -0700" }, "message": "nvme: set queue limits for the admin queue\n\nFactor out a helper to set all the device specific queue limits and apply\nthem to the admin queue in addition to the I/O queues. Without this the\ncommand size on the admin queue is arbitrarily low, and the missing\nother limitations are just minefields waiting for victims.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nReported-by: Jeff Lien \u003cJeff.Lien@hgst.com\u003e\nTested-by: Jeff Lien \u003cJeff.Lien@hgst.com\u003e\nReviewed-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "a1a0e23e49037c23ea84bc8cc146a03584d13577", "tree": "b8bb7792ed9dd97358bdea0a6a2f36fdb617bbf0", "parents": [ "e9fc63d682dbbef17921aeb00d03fd52d6735ffd" ], "author": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Mon Feb 29 18:28:53 2016 -0500" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:50 2016 -0700" }, "message": "writeback: flush inode cgroup wb switches instead of pinning super_block\n\nIf cgroup writeback is in use, inodes can be scheduled for\nasynchronous wb switching. Before 5ff8eaac1636 (\"writeback: keep\nsuperblock pinned during cgroup writeback association switches\"), this\ncould race with umount leading to super_block being destroyed while\ninodes are pinned for wb switching. 5ff8eaac1636 fixed it by bumping\ns_active while wb switches are in flight; however, this allowed\nin-flight wb switches to make umounts asynchronous when the userland\nexpected synchronosity - e.g. fsck immediately following umount may\nfail because the device is still busy.\n\nThis patch removes the problematic super_block pinning and instead\nmakes generic_shutdown_super() flush in-flight wb switches. wb\nswitches are now executed on a dedicated isw_wq so that they can be\nflushed and isw_nr_in_flight keeps track of the number of in-flight wb\nswitches so that flushing can be avoided in most cases.\n\nv2: Move cgroup_writeback_umount() further below and add MS_ACTIVE\n check in inode_switch_wbs() as Jan an Al suggested.\n\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\nReported-by: Tahsin Erdogan \u003ctahsin@google.com\u003e\nCc: Jan Kara \u003cjack@suse.cz\u003e\nCc: Al Viro \u003cviro@ZenIV.linux.org.uk\u003e\nLink: http://lkml.kernel.org/g/CAAeU0aNCq7LGODvVGRU-oU_o-6enii5ey0p1c26D1ZzYwkDc5A@mail.gmail.com\nFixes: 5ff8eaac1636 (\"writeback: keep superblock pinned during cgroup writeback association switches\")\nCc: stable@vger.kernel.org #v4.5\nReviewed-by: Jan Kara \u003cjack@suse.cz\u003e\nTested-by: Tahsin Erdogan \u003ctahsin@google.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "e9fc63d682dbbef17921aeb00d03fd52d6735ffd", "tree": "33d5f3ece18a1724ad42b79a7c4970859efca5e8", "parents": [ "63088ec7c8eadfe08b96127a41b385ec9742dace" ], "author": { "name": "Keith Busch", "email": "keith.busch@intel.com", "time": "Wed Feb 24 09:15:58 2016 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:50 2016 -0700" }, "message": "NVMe: Fix 0-length integrity payload\n\nA user could send a passthrough IO command with a metadata pointer to a\nnamespace without metadata. With metadata length of 0, kmalloc returns\nZERO_SIZE_PTR. Since that is not NULL, the driver would have set this as\nthe bio\u0027s integrity payload, which causes an access fault on completion.\n\nThis patch ignores the users metadata buffer if the namespace format\ndoes not support separate metadata.\n\nReported-by: Stephen Bates \u003cstephen.bates@microsemi.com\u003e\nSigned-off-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Sagi Grimberg \u003csagig@mellanox.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Johannes Thumshirn \u003cjthumshirn@suse.de\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "63088ec7c8eadfe08b96127a41b385ec9742dace", "tree": "bf9ced59312fab61bec342bb406b2367c74040b1", "parents": [ "69d9a99c258eb1d6478fd9608a2070890797eed7" ], "author": { "name": "Keith Busch", "email": "keith.busch@intel.com", "time": "Wed Feb 24 09:15:57 2016 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:50 2016 -0700" }, "message": "NVMe: Don\u0027t allow unsupported flags\n\nThe command flags can change the meaning of other fields in the command\nthat the driver is not prepared to handle. Specifically, the user could\npassthrough an SGL flag, causing the controller to misinterpret the PRP\nlist the driver created, potentially corrupting memory or data.\n\nSigned-off-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Jon Derrick \u003cjonathan.derrick@intel.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Johannes Thumshirn \u003cjthumshirn@suse.de\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "69d9a99c258eb1d6478fd9608a2070890797eed7", "tree": "c431aebc0a314c5f84604f4bf609482df5248938", "parents": [ "f58944e265d4ebe47216a5d7488aee3928823d30" ], "author": { "name": "Keith Busch", "email": "keith.busch@intel.com", "time": "Wed Feb 24 09:15:56 2016 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:50 2016 -0700" }, "message": "NVMe: Move error handling to failed reset handler\n\nThis moves failed queue handling out of the namespace removal path and\ninto the reset failure path, fixing a hanging condition if the controller\nfails or link down during del_gendisk. Previously the driver had to see\nthe controller as degraded prior to calling del_gendisk to setup the\nqueues to fail. But, if the controller happened to fail after this,\nthere was no task to end outstanding requests.\n\nOn failure, all namespace states are set to dead. This has capacity\nrevalidate to 0, and ends all new requests with error status.\n\nSigned-off-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Johannes Thumshirn \u003cjthumshirn@suse.de\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "f58944e265d4ebe47216a5d7488aee3928823d30", "tree": "57f2e0e81b2df9df68dd7d3223212570eef04474", "parents": [ "646017a612e72f19bd9f991fe25287a149c5f627" ], "author": { "name": "Keith Busch", "email": "keith.busch@intel.com", "time": "Wed Feb 24 09:15:55 2016 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:49 2016 -0700" }, "message": "NVMe: Simplify device reset failure\n\nA reset failure schedules the device to unbind from the driver through\nthe pci driver\u0027s remove. This cleans up all intialization, so there is\nno need to duplicate the potentially racy cleanup.\n\nTo help understand why a reset failed, the status is logged with the\nexisting warning message.\n\nSigned-off-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Johannes Thumshirn \u003cjthumshirn@suse.de\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "646017a612e72f19bd9f991fe25287a149c5f627", "tree": "c8ff5e127118a4b6de68734be5a8e0560110debe", "parents": [ "075790ebba4a1eb297f9875e581b55c0382b1f3d" ], "author": { "name": "Keith Busch", "email": "keith.busch@intel.com", "time": "Wed Feb 24 09:15:54 2016 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:49 2016 -0700" }, "message": "NVMe: Fix namespace removal deadlock\n\nThis patch makes nvme namespace removal lockless. It is up to the caller\nto ensure no active namespace scanning is occuring. To ensure no scan\nwork occurs, the nvme pci driver adds a removing state to the controller\ndevice to avoid queueing scan work during removal. The work is flushed\nafter setting the state, so no new scan work can be queued.\n\nThe lockless removal allows the driver to cleanup a namespace\nrequest_queue if the controller fails during removal. Previously this\ncould deadlock trying to acquire the namespace mutex in order to handle\nsuch events.\n\nSigned-off-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Johannes Thumshirn \u003cjthumshirn@suse.de\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "075790ebba4a1eb297f9875e581b55c0382b1f3d", "tree": "0b6119c4f7decc6b2809be19006b951e88e34f17", "parents": [ "b00a726a9fd82ddd4c10344e46f0d371e1674303" ], "author": { "name": "Keith Busch", "email": "keith.busch@intel.com", "time": "Wed Feb 24 09:15:53 2016 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:49 2016 -0700" }, "message": "NVMe: Use IDA for namespace disk naming\n\nA namespace may be detached from a controller, but a user may be holding\na reference to it. Attaching a new namespace with the same NSID will create\nduplicate names when using the NSID to name the disk.\n\nThis patch uses an IDA that is released only when the last reference is\nreleased instead of using the namespace ID.\n\nSigned-off-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Sagi Grimberg \u003csagig@mellanox.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nReviewed-by: Johannes Thumshirn \u003cjthumshirn@suse.de\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "b00a726a9fd82ddd4c10344e46f0d371e1674303", "tree": "c70b623e3d4a81b2477dc36103a316fef3a1b440", "parents": [ "e827091cb1bcd8e718ac3657845fb809c0b93324" ], "author": { "name": "Keith Busch", "email": "keith.busch@intel.com", "time": "Wed Feb 24 09:15:52 2016 -0700" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:49 2016 -0700" }, "message": "NVMe: Don\u0027t unmap controller registers on reset\n\nUnmapping the registers on reset or shutdown is not necessary. Keeping\nthe mapping simplifies reset handling.\n\nSigned-off-by: Keith Busch \u003ckeith.busch@intel.com\u003e\nReviewed-by: Johannes Thumshirn \u003cjthumshirn@suse.de\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "e827091cb1bcd8e718ac3657845fb809c0b93324", "tree": "146811f42b518408c3ead1c78c51a3198d811f61", "parents": [ "25e71a99f10e444cd00bb2ebccb11e1c9fb672b1" ], "author": { "name": "Ming Lei", "email": "ming.lei@canonical.com", "time": "Fri Feb 26 23:40:53 2016 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:49 2016 -0700" }, "message": "block: merge: get the 1st and last bvec via helpers\n\nThis patch applies the two introduced helpers to\nfigure out the 1st and last bvec.\n\nReviewed-by: Sagi Grimberg \u003csagig@mellanox.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Ming Lei \u003cming.lei@canonical.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "25e71a99f10e444cd00bb2ebccb11e1c9fb672b1", "tree": "70219779b021ec7cd2d32bc1091208a126fe0bec", "parents": [ "e0af29171aa8912e1ca95023b75ef336cd70d661" ], "author": { "name": "Ming Lei", "email": "ming.lei@canonical.com", "time": "Fri Feb 26 23:40:52 2016 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:49 2016 -0700" }, "message": "block: get the 1st and last bvec via helpers\n\nThis patch applies the two introduced helpers to\nfigure out the 1st and last bvec, and fixes the\noriginal way after bio splitting.\n\nCc: stable@vger.kernel.org\nReported-by: Sagi Grimberg \u003csagig@dev.mellanox.co.il\u003e\nReviewed-by: Sagi Grimberg \u003csagig@mellanox.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Ming Lei \u003cming.lei@canonical.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "e0af29171aa8912e1ca95023b75ef336cd70d661", "tree": "0e8bd6fdfb2b96eb505f71f306afbab72d4f3c88", "parents": [ "7bcd79ac50d9d83350a835bdb91c04ac9e098412" ], "author": { "name": "Ming Lei", "email": "ming.lei@canonical.com", "time": "Fri Feb 26 23:40:51 2016 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:49 2016 -0700" }, "message": "block: check virt boundary in bio_will_gap()\n\nIn the following patch, the way for figuring out\nthe last bvec will be changed with a bit cost introduced,\nso return immediately if the queue doesn\u0027t have virt\nboundary limit. Actually most of devices have not\nthis limit.\n\nReviewed-by: Sagi Grimberg \u003csagig@mellanox.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Ming Lei \u003cming.lei@canonical.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "7bcd79ac50d9d83350a835bdb91c04ac9e098412", "tree": "b8161854b8ec9825e7b9fabf2299231130eaf574", "parents": [ "e3c2ef41f88e50c8557270868600d3132028af3b" ], "author": { "name": "Ming Lei", "email": "ming.lei@canonical.com", "time": "Fri Feb 26 23:40:50 2016 +0800" }, "committer": { "name": "Jens Axboe", "email": "axboe@fb.com", "time": "Thu Mar 03 14:42:49 2016 -0700" }, "message": "block: bio: introduce helpers to get the 1st and last bvec\n\nThe bio passed to bio_will_gap() may be fast cloned from upper\nlayer(dm, md, bcache, fs, ...), or from bio splitting in block\ncore.\n\nUnfortunately bio_will_gap() just figures out the last bvec via\n\u0027bi_io_vec[prev-\u003ebi_vcnt - 1]\u0027 directly, and this way is obviously\nwrong.\n\nThis patch introduces two helpers for getting the first and last\nbvec of one bio for fixing the issue.\n\nCc: stable@vger.kernel.org\nReported-by: Sagi Grimberg \u003csagig@dev.mellanox.co.il\u003e\nReviewed-by: Sagi Grimberg \u003csagig@mellanox.com\u003e\nReviewed-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Ming Lei \u003cming.lei@canonical.com\u003e\nSigned-off-by: Jens Axboe \u003caxboe@fb.com\u003e\n" }, { "commit": "fbe093ac9f0201939279cdfe8b0fce20ce5ef7a9", "tree": "c1013fbe8e968d678d26867e7cef9279a42e816a", "parents": [ "93125094c07d8c9ec25dff5869f191b33eb9dd6e" ], "author": { "name": "Sakari Ailus", "email": "sakari.ailus@iki.fi", "time": "Thu Mar 03 14:20:14 2016 -0300" }, "committer": { "name": "Mauro Carvalho Chehab", "email": "mchehab@osg.samsung.com", "time": "Thu Mar 03 18:10:53 2016 -0300" }, "message": "[media] media: Sanitise the reserved fields of the G_TOPOLOGY IOCTL arguments\n\nThe argument structs are used in arrays for G_TOPOLOGY IOCTL. The\narguments themselves do not need to be aligned to a power of two, but\naligning them up to the largest basic type alignment (u64) on common ABIs\nis a good thing to do.\n\nThe patch changes the size of the reserved fields to 5 or 6 u32\u0027s and\naligns the size of the struct to 8 bytes so we do no longer depend on the\ncompiler to perform the alignment.\n\nWhile at it, add __attribute__ ((packed)) to these structs as well.\n\nSigned-off-by: Sakari Ailus \u003csakari.ailus@linux.intel.com\u003e\nSigned-off-by: Mauro Carvalho Chehab \u003cmchehab@osg.samsung.com\u003e\n" }, { "commit": "e3c2ef41f88e50c8557270868600d3132028af3b", "tree": "7a452fdc5ac2d0ffb72ef34a86636b68ca951853", "parents": [ "c2687cf9502a4b69abc0f3a7b1618ffef47a40db", "1195c103f6c98d9ff381cac3a8760d4f8a133627" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 12:54:39 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 12:54:39 2016 -0800" }, "message": "Merge tag \u0027pci-v4.5-fixes-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci\n\nPull PCI fixes from Bjorn Helgaas:\n \"Freescale Layerscape host bridge driver:\n Fix MSG TLP drop setting (Minghuan Lian)\n\n TI Keystone host bridge driver:\n Fix MSI code that retrieves struct pcie_port pointer (Murali Karicheri)\"\n\n* tag \u0027pci-v4.5-fixes-4\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:\n PCI: layerscape: Fix MSG TLP drop setting\n PCI: keystone: Fix MSI code that retrieves struct pcie_port pointer\n" }, { "commit": "861c3849222b7cadfd0da63f9c7573c2f15ba535", "tree": "6ff618044a493bf2f03559b3de7c5173d4c2ca84", "parents": [ "f4f41c5abcf9acc61939367dc80fd62d19b16edb", "04fdbc825ffc02fb098964b92de802fff44e73fd" ], "author": { "name": "Greg Kroah-Hartman", "email": "gregkh@linuxfoundation.org", "time": "Thu Mar 03 12:37:21 2016 -0800" }, "committer": { "name": "Greg Kroah-Hartman", "email": "gregkh@linuxfoundation.org", "time": "Thu Mar 03 12:37:21 2016 -0800" }, "message": "Merge tag \u0027usb-serial-4.5-rc7\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/johan/usb-serial into usb-linus\n\nJonan writes:\n\nUSB-serial fixes for v4.5-rc7\n\nHere are some new device ids and a patch removing the mxu11x0 driver,\nwhich turned out not to be needed.\n\nSigned-off-by: Johan Hovold \u003cjohan@kernel.org\u003e\n" }, { "commit": "c2687cf9502a4b69abc0f3a7b1618ffef47a40db", "tree": "2af1ab08b71c349fe27225fa059cebfac38553ab", "parents": [ "4237b2e6c6ca77b72f0b45ffad29e284e776ede3", "2680d6da455b636dd006636780c0f235c6561d70" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 11:54:56 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 11:54:56 2016 -0800" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm\n\nPull KVM fixes from Paolo Bonzini:\n - ARM/MIPS: Fixes for ioctls when copy_from_user returns nonzero\n - x86: Small fix for Skylake TSC scaling\n - x86: Improved fix for last week\u0027s missed hardware breakpoint bug\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/virt/kvm/kvm:\n kvm: x86: Update tsc multiplier on change.\n mips/kvm: fix ioctl error handling\n arm/arm64: KVM: Fix ioctl error handling\n KVM: x86: fix root cause for missed hardware breakpoints\n" }, { "commit": "4237b2e6c6ca77b72f0b45ffad29e284e776ede3", "tree": "5f4c0b52305850928301d167990dc7d2d72d159e", "parents": [ "19eab220e7a00bbb36543b4daf5e3c464ec474cb", "b26a719bdba9aa926ceaadecc66e07623d2b8a53" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 11:39:11 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 11:39:11 2016 -0800" }, "message": "Merge tag \u0027gpio-v4.5-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio\n\nPull late GPIO fix from Linus Walleij:\n \"Regressions never arrive when you want them to, so here is a late fix\n for the Renesas RCAR GPIO driver. It only affects that driver on the\n very specific Renesas platforms:\n\n - Fix a runtime PM suspend/resume bug in the RCAR driver\"\n\n* tag \u0027gpio-v4.5-3\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio:\n gpio: rcar: Add Runtime PM handling for interrupts\n" }, { "commit": "19eab220e7a00bbb36543b4daf5e3c464ec474cb", "tree": "a1e29c3dd6d5a35850882e3b9ab149da29c1c7f1", "parents": [ "f4bd9822080ece335663245c8520b146d9e1bdad", "e6a8c9b337eed56eb481e1b4dd2180c25a1e5310" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 11:32:13 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 11:32:13 2016 -0800" }, "message": "Merge tag \u0027iommu-fixes-v4.5-rc6\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu\n\nPull IOMMU fixes from Joerg Roedel:\n \"One fix for Intel VT-d:\n\n - Use BUS_NOTIFY_REMOVED_DEVICE notifier to unbind a device from its\n domain _after_ it has been unbound from its driver. This fixes a\n BUG_ON being triggered in the PCI hotplug path.\n\n And three for AMD IOMMU:\n\n - Add a workaround for a hardware issue with ATS in use\n\n - Fix ATS enable/disable balance when a device is removed\n\n - Fix a boot warning being triggered when the system has IOMMU\n performance counters and PCI device 00:00.0 is not covered by the\n IOMMU\"\n\n* tag \u0027iommu-fixes-v4.5-rc6\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/joro/iommu:\n iommu/vt-d: Use BUS_NOTIFY_REMOVED_DEVICE in hotplug path\n iommu/amd: Detach device from domain before removal\n iommu/amd: Apply workaround for ATS write permission check\n iommu/amd: Fix boot warning when device 00:00.0 is not iommu covered\n" }, { "commit": "f4bd9822080ece335663245c8520b146d9e1bdad", "tree": "5718c3b84741c97b1aba4d777a9ef88f37a9a6fd", "parents": [ "52ad12966b0217a6dcf782cc2516061343911c0d", "e1f33be9186363da7955bcb5f0b03e6685544c50" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 11:07:32 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 11:07:32 2016 -0800" }, "message": "Merge tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost\n\nPull minor virtio/vhost fixes from Michael Tsirkin:\n \"This fixes two minor bugs: error handling in vhost, and capability\n processing in virtio\"\n\n* tag \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost:\n vhost: fix error path in vhost_init_used()\n virtio-pci: read the right virtio_pci_notify_cap field\n" }, { "commit": "52ad12966b0217a6dcf782cc2516061343911c0d", "tree": "9684a7e361be5c05fd65e3e0c389bbbd3d2c85b2", "parents": [ "215b031b970a6cec729dcfaa94537074256ba2e8", "8160c4e455820d5008a1116d2dca35f0363bb062" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 10:56:17 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 10:56:17 2016 -0800" }, "message": "Merge tag \u0027vfio-v4.5-rc7\u0027 of git://github.com/awilliam/linux-vfio\n\nPull VFIO fix from Alex Williamson:\n \"Use -EFAULT for copy_to_user error in ioctl (Michael Tsirkin)\"\n\n* tag \u0027vfio-v4.5-rc7\u0027 of git://github.com/awilliam/linux-vfio:\n vfio: fix ioctl error handling\n" }, { "commit": "215b031b970a6cec729dcfaa94537074256ba2e8", "tree": "a79e414ea703174d5f300aa498dfdc81631a3a3a", "parents": [ "f983cd32cd5d2ec47ac8868f387bad15e8f11f85", "a1e533ec07d583d01349ef13c0c965b8633e1b91" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 10:46:18 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Mar 03 10:46:18 2016 -0800" }, "message": "Merge tag \u0027fbdev-fixes-4.5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tomba/linux\n\nPull fbdev fix from Tomi Valkeinen:\n \"Fix hang caused by fbconsole blink timer\"\n\n* tag \u0027fbdev-fixes-4.5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tomba/linux:\n fbcon: set a default value to blink interval\n" }, { "commit": "93125094c07d8c9ec25dff5869f191b33eb9dd6e", "tree": "ddb760a612ee0633681543d0ccf38a03e361c95a", "parents": [ "58402b6e98f9059420e64b296db6c4cb6f6c1117" ], "author": { "name": "Mauro Carvalho Chehab", "email": "mchehab@osg.samsung.com", "time": "Thu Mar 03 14:52:51 2016 -0300" }, "committer": { "name": "Mauro Carvalho Chehab", "email": "mchehab@osg.samsung.com", "time": "Thu Mar 03 14:52:51 2016 -0300" }, "message": "[media] media.h: postpone connectors entities\n\nThe representation of external connections got some heated\ndiscussions recently. As we\u0027re too close to the merge window,\nlet\u0027s not set those entities into a stone.\n\nSigned-off-by: Mauro Carvalho Chehab \u003cmchehab@osg.samsung.com\u003e\n" }, { "commit": "11d8d645343efba0c975aefe7c2cf3b33c836c75", "tree": "ee2586a426241d0f1f9e4e9a94572d6c420a0d07", "parents": [ "3d943c9d1cc5ad1825e46291ef5ce627e1b6b660" ], "author": { "name": "Or Gerlitz", "email": "ogerlitz@mellanox.com", "time": "Tue Mar 01 18:52:23 2016 +0200" }, "committer": { "name": "Doug Ledford", "email": "dledford@redhat.com", "time": "Thu Mar 03 10:52:58 2016 -0500" }, "message": "IB/core: Use GRH when the path hop-limit \u003e 0\n\nAccording to IBTA spec v1.3 section 12.7.19, QPs should use GRH when\nthe path returned by the SA has hop-limit \u003e 0. Currently, we do that\nonly for the \u003e 1 case, fix that.\n\nFixes: 6d969a471ba1 (\u0027IB/sa: Add ib_init_ah_from_path()\u0027)\nSigned-off-by: Or Gerlitz \u003cogerlitz@mellanox.com\u003e\nReviewed-by: Jason Gunthorpe \u003cjgunthorpe@obsidianresearch.com\u003e\nSigned-off-by: Doug Ledford \u003cdledford@redhat.com\u003e\n" }, { "commit": "f16921275cc3c2442d0b95225785a601603b990f", "tree": "712526e22bfe1e38ce5615f5c0530a47c03ccd68", "parents": [ "81f70ba233d5f660e1ea5fe23260ee323af5d53a" ], "author": { "name": "Robert Jarzmik", "email": "robert.jarzmik@free.fr", "time": "Tue Feb 16 22:54:02 2016 +0100" }, "committer": { "name": "Vinod Koul", "email": "vinod.koul@intel.com", "time": "Thu Mar 03 21:06:45 2016 +0530" }, "message": "dmaengine: pxa_dma: fix cyclic transfers\n\nWhile testing audio with pxa2xx-ac97, underrun were happening while the\nuser application was correctly feeding the music. Debug proved that the\ncyclic transfer is not cyclic, ie. the last descriptor did not loop on\nthe first.\n\nAnother issue is that the descriptor length was always set to 8192,\nbecause of an trivial operator issue.\n\nThis was tested on a pxa27x platform.\n\nFixes: a57e16cf0333 (\"dmaengine: pxa: add pxa dmaengine driver\")\nReported-by: Vasily Khoruzhick \u003canarsoul@gmail.com\u003e\nTested-by: Vasily Khoruzhick \u003canarsoul@gmail.com\u003e\nSigned-off-by: Robert Jarzmik \u003crobert.jarzmik@free.fr\u003e\nSigned-off-by: Vinod Koul \u003cvinod.koul@intel.com\u003e\n" }, { "commit": "3d943c9d1cc5ad1825e46291ef5ce627e1b6b660", "tree": "1e8382403949388ffede81a275dc5c81bb3a6072", "parents": [ "85d9691ccc96d95629939a877fd6c1f8c4724f56" ], "author": { "name": "Majd Dibbiny", "email": "majd@mellanox.com", "time": "Sun Feb 14 18:35:52 2016 +0200" }, "committer": { "name": "Doug Ledford", "email": "dledford@redhat.com", "time": "Thu Mar 03 10:00:18 2016 -0500" }, "message": "IB/{core, mlx5}: Fix input len in vendor part of create_qp/srq\n\nCurrently, the inlen field of the vendor\u0027s part of the command\ndoesn\u0027t match the command buffer. This happens because the inlen\naccommodates ib_uverbs_cmd_hdr which is deducted from the in buffer.\nThis is problematic since the vendor function could be called either\nfrom the legacy verb (where the input length mismatches the actual\nlength) or by the extended verb (where the length matches). The vendor\nhas no idea which function calls it and therefore has no way to know\nhow the length variable should be treated.\n\nFixing this by aligning the inlen to the correct length.\n\nAll vendor drivers either assumed that inlen \u003e\u003d sizeof(vendor_uhw_cmd)\nor just failed wrongly (mlx5) and fixed in this patch.\n\nFixes: cfb5e088e26a (\u0027IB/mlx5: Add CQE version 1 support to user QPs and SRQs\u0027)\nSigned-off-by: Majd Dibbiny \u003cmajd@mellanox.com\u003e\nReviewed-by: Matan Barak \u003cmatanb@mellanox.com\u003e\nReviewed-by: Haggai Eran \u003chaggaie@mellanox.com\u003e\nSigned-off-by: Doug Ledford \u003cdledford@redhat.com\u003e\n" }, { "commit": "85d9691ccc96d95629939a877fd6c1f8c4724f56", "tree": "03325bdfe51f55708b61046de40e5275f278a6dd", "parents": [ "5adebafb75bdfbbe4ec69f14c3613e70f6ed7f6f" ], "author": { "name": "Majd Dibbiny", "email": "majd@mellanox.com", "time": "Sun Feb 14 18:35:51 2016 +0200" }, "committer": { "name": "Doug Ledford", "email": "dledford@redhat.com", "time": "Thu Mar 03 10:00:03 2016 -0500" }, "message": "IB/mlx5: Avoid using user-index for SRQs\n\nNormal SRQs, unlike XRC SRQs, don\u0027t have user-index, therefore\navoid verifying it and using it.\n\nFixes: cfb5e088e26a (\u0027IB/mlx5: Add CQE version 1 support to user QPs and SRQs\u0027)\nSigned-off-by: Majd Dibbiny \u003cmajd@mellanox.com\u003e\nReviewed-by: Matan Barak \u003cmatanb@mellanox.com\u003e\nSigned-off-by: Doug Ledford \u003cdledford@redhat.com\u003e\n" }, { "commit": "fb822e6076d972691c5dd33431c10f82dda94ae9", "tree": "95aefc3d511923fdb5438166bbb4d7e70cddfa85", "parents": [ "923adb1646d5ba739d2a1e63ee20d60574d9da8e" ], "author": { "name": "Ravi Bangoria", "email": "ravi.bangoria@linux.vnet.ibm.com", "time": "Wed Mar 02 15:25:17 2016 +0530" }, "committer": { "name": "Michael Ellerman", "email": "mpe@ellerman.id.au", "time": "Thu Mar 03 22:06:08 2016 +1100" }, "message": "powerpc/hw_breakpoint: Fix oops when destroying hw_breakpoint event\n\nWhen destroying a hw_breakpoint event, the kernel oopses as follows:\n\n Unable to handle kernel paging request for data at address 0x00000c07\n NIP [c0000000000291d0] arch_unregister_hw_breakpoint+0x40/0x60\n LR [c00000000020b6b4] release_bp_slot+0x44/0x80\n\nCall chain:\n\n hw_breakpoint_event_init()\n bp-\u003edestroy \u003d bp_perf_event_destroy;\n\n do_exit()\n perf_event_exit_task()\n perf_event_exit_task_context()\n WRITE_ONCE(child_ctx-\u003etask, TASK_TOMBSTONE);\n perf_event_exit_event()\n free_event()\n _free_event()\n bp_perf_event_destroy() // event-\u003edestroy(event);\n release_bp_slot()\n arch_unregister_hw_breakpoint()\n\nperf_event_exit_task_context() sets child_ctx-\u003etask as TASK_TOMBSTONE\nwhich is (void *)-1. arch_unregister_hw_breakpoint() tries to fetch\n\u0027thread\u0027 attribute of \u0027task\u0027 resulting in oops.\n\nPeterz points out that the code shouldn\u0027t be using bp-\u003ectx anyway, but\nfixing that will require a decent amount of rework. So for now to fix\nthe oops, check if bp-\u003ectx-\u003etask has been set to (void *)-1, before\ndereferencing it. We don\u0027t use TASK_TOMBSTONE, because that would\nrequire exporting it and it\u0027s supposed to be an internal detail.\n\nFixes: 63b6da39bb38 (\"perf: Fix perf_event_exit_task() race\")\nSigned-off-by: Ravi Bangoria \u003cravi.bangoria@linux.vnet.ibm.com\u003e\nAcked-by: Peter Zijlstra (Intel) \u003cpeterz@infradead.org\u003e\nSigned-off-by: Michael Ellerman \u003cmpe@ellerman.id.au\u003e\n" }, { "commit": "02322ac9dee9aff8d8862e8d6660ebe102f492ea", "tree": "5098a90d69bac7d4806ca81534cfc248e2baa454", "parents": [ "197b958c1e76a575d77038cc98b4bebc2134279f" ], "author": { "name": "Simon South", "email": "simon@simonsouth.com", "time": "Wed Mar 02 23:10:44 2016 -0500" }, "committer": { "name": "Takashi Iwai", "email": "tiwai@suse.de", "time": "Thu Mar 03 10:27:54 2016 +0100" }, "message": "ALSA: hda - Fix mic issues on Acer Aspire E1-472\n\nThis patch applies the microphone-related fix created for the Acer\nAspire E1-572 to the E1-472 as well, as it uses the same Realtek ALC282\nCODEC and demonstrates the same issues.\n\nThis patch allows an external, headset microphone to be used and limits\nthe gain on the (quite noisy) internal microphone.\n\nSigned-off-by: Simon South \u003csimon@simonsouth.com\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Takashi Iwai \u003ctiwai@suse.de\u003e\n" }, { "commit": "58402b6e98f9059420e64b296db6c4cb6f6c1117", "tree": "f223f14983b4f395aad3f112ee42afc936ca29df", "parents": [ "0ba4581c84cfb39fd527f6b3457f1c97f6356c04" ], "author": { "name": "Hans Verkuil", "email": "hans.verkuil@cisco.com", "time": "Mon Feb 29 09:02:47 2016 +0100" }, "committer": { "name": "Mauro Carvalho Chehab", "email": "mchehab@osg.samsung.com", "time": "Thu Mar 03 05:56:33 2016 -0300" }, "message": "[media] media.h: use hex values for range offsets, move connectors base up.\n\nMake the base offset hexadecimal to simplify debugging since the base\naddresses are hex too.\n\nThe offsets for connectors is also changed to start after the \u0027reserved\u0027\nrange 0x10000-0x2ffff.\n\nSigned-off-by: Hans Verkuil \u003chans.verkuil@cisco.com\u003e\nSigned-off-by: Mauro Carvalho Chehab \u003cmchehab@osg.samsung.com\u003e\n" }, { "commit": "f0511e66114a6414cfca92d7b94118913a0c11ff", "tree": "540c2120b02dd1717e8ae5e6eddc8520c49e8cf8", "parents": [ "2d02b8bdba322b527c5f5168ce1ca10c2d982a78", "0b39c531cfa12dad54eac238c2e303b994df1ef7" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Mar 03 11:37:07 2016 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Thu Mar 03 11:37:07 2016 +1000" }, "message": "Merge branch \u0027drm-fixes-4.5\u0027 of git://people.freedesktop.org/~agd5f/linux into drm-fixes\n\nFixes for radeon and amdgpu:\n- Fix GPUVM flushing on CI and VI\n- Misc DPM and Powerplay fixes\n- VCE DPM fixes for CZ/ST\n- DP hotplug fix\n\n* \u0027drm-fixes-4.5\u0027 of git://people.freedesktop.org/~agd5f/linux:\n drm/amdgpu: return from atombios_dp_get_dpcd only when error\n drm/amdgpu/cz: remove commented out call to enable vce pg\n drm/amdgpu/powerplay/cz: enable/disable vce dpm independent of vce pg\n drm/amdgpu/cz: enable/disable vce dpm even if vce pg is disabled\n drm/amdgpu/gfx8: specify which engine to wait before vm flush\n drm/amdgpu: apply gfx_v8 fixes to gfx_v7 as well\n drm/amd/powerplay: send event to notify powerplay all modules are initialized.\n drm/amd/powerplay: export AMD_PP_EVENT_COMPLETE_INIT task to amdgpu.\n drm/radeon/pm: update current crtc info after setting the powerstate\n drm/amdgpu/pm: update current crtc info after setting the powerstate\n" }, { "commit": "92f9e179a702a6adbc11e2fedc76ecd6ffc9e3f7", "tree": "72ee540a52c68f71d4cf7b13bc8fa188bee8759f", "parents": [ "fc77dbd34c5c99bce46d40a2491937c3bcbd10af" ], "author": { "name": "Todd E Brandt", "email": "todd.e.brandt@linux.intel.com", "time": "Wed Mar 02 16:05:29 2016 -0800" }, "committer": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Thu Mar 03 02:28:28 2016 +0100" }, "message": "PM / sleep / x86: Fix crash on graph trace through x86 suspend\n\nPause/unpause graph tracing around do_suspend_lowlevel as it has\ninconsistent call/return info after it jumps to the wakeup vector.\nThe graph trace buffer will otherwise become misaligned and\nmay eventually crash and hang on suspend.\n\nTo reproduce the issue and test the fix:\nRun a function_graph trace over suspend/resume and set the graph\nfunction to suspend_devices_and_enter. This consistently hangs the\nsystem without this fix.\n\nSigned-off-by: Todd Brandt \u003ctodd.e.brandt@linux.intel.com\u003e\nCc: All applicable \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\n" }, { "commit": "f3c87e99f401a0ff0ef941346ec6e790f0fd526e", "tree": "8a68bf6a041ded79b098bd71999bf28c5140e36d", "parents": [ "fc77dbd34c5c99bce46d40a2491937c3bcbd10af", "949024d6702d1da8a266f42226d8402f0df078dc" ], "author": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Wed Mar 02 23:24:33 2016 +0100" }, "committer": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Wed Mar 02 23:24:33 2016 +0100" }, "message": "Merge tag \u0027renesas-dt-fixes2-for-v4.5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas into fixes\n\nMerge \"Second Round of Renesas ARM Based SoC DT Fixes for v4.5\" from Simon Horman:\n\n* remove enable prop from HS-USB device node on porter board\n\n* tag \u0027renesas-dt-fixes2-for-v4.5\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/horms/renesas:\n ARM: dts: porter: remove enable prop from HS-USB device node\n" }, { "commit": "f983cd32cd5d2ec47ac8868f387bad15e8f11f85", "tree": "535cf2325fafd13c794d6d0dc2599e7191297d3c", "parents": [ "12f1d7e4932a40100177be5681f1a26f46265217", "b4f09ae6db2a98ff7817c3371aaab39ca22daf80" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 02 09:46:19 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 02 09:46:19 2016 -0800" }, "message": "Merge branch \u0027parisc-4.5-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux\n\nPull parisc fixes from Helge Deller:\n \"We wire up the copy_file_range syscall, fix two bugs in the parisc\n ptrace code and have a trivial fix for floppy.h to clarify an\n expression with parentheses\"\n\n* \u0027parisc-4.5-2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/deller/parisc-linux:\n parisc: Wire up copy_file_range syscall\n parisc: Fix ptrace syscall number and return value modification\n parisc: Use parentheses around expression in floppy.h\n" }, { "commit": "12f1d7e4932a40100177be5681f1a26f46265217", "tree": "e8e1795663ae7d227bf4490d2876968fa7ca0c43", "parents": [ "39680f50ae54cbbb6e72ac38b8329dd3eb9105f4", "9589995e46d89c70c6fa4723c5f3e5ec04c3c3e3" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 02 09:15:21 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 02 09:15:21 2016 -0800" }, "message": "Merge branch \u0027for-next\u0027 of git://git.samba.org/sfrench/cifs-2.6\n\nPull cifs fixes from Steve French:\n \"Various small CIFS/SMB3 fixes for stable:\n\n Fixes address oops that can occur when accessing Macs with SMB3, and\n another problem found to Samba when read responses queued (e.g. with\n gluster under Samba)\"\n\n* \u0027for-next\u0027 of git://git.samba.org/sfrench/cifs-2.6:\n CIFS: Fix duplicate line introduced by clone_file_range patch\n Fix cifs_uniqueid_to_ino_t() function for s390x\n CIFS: Fix SMB2+ interim response processing for read requests\n cifs: fix out-of-bounds access in lease parsing\n" }, { "commit": "39680f50ae54cbbb6e72ac38b8329dd3eb9105f4", "tree": "b8eeff48b7a5d0aa733a663cd379455037c13769", "parents": [ "f691b77b1fc491dae601631c8531a0a13e915466" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Mar 01 11:56:22 2016 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Mar 02 09:03:18 2016 -0800" }, "message": "userfaultfd: don\u0027t block on the last VM updates at exit time\n\nThe exit path will do some final updates to the VM of an exiting process\nto inform others of the fact that the process is going away.\n\nThat happens, for example, for robust futex state cleanup, but also if\nthe parent has asked for a TID update when the process exits (we clear\nthe child tid field in user space).\n\nHowever, at the time we do those final VM accesses, we\u0027ve already\nstopped accepting signals, so the usual \"stop waiting for userfaults on\nsignal\" code in fs/userfaultfd.c no longer works, and the process can\nbecome an unkillable zombie waiting for something that will never\nhappen.\n\nTo solve this, just make handle_userfault() abort any user fault\nhandling if we\u0027re already in the exit path past the signal handling\nstate being dead (marked by PF_EXITING).\n\nThis VM special case is pretty ugly, and it is possible that we should\nlook at finalizing signals later (or move the VM final accesses\nearlier). But in the meantime this is a fairly minimally intrusive fix.\n\nReported-and-tested-by: Dmitry Vyukov \u003cdvyukov@google.com\u003e\nAcked-by: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "0b39c531cfa12dad54eac238c2e303b994df1ef7", "tree": "ced972693ee3a4878fe174d3319202a3f35a2e8b", "parents": [ "89913ea615754163ae9a066c9c1b95aa2a2f51b4" ], "author": { "name": "Arindam Nath", "email": "arindam.nath@amd.com", "time": "Wed Mar 02 17:19:01 2016 +0530" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:25 2016 -0500" }, "message": "drm/amdgpu: return from atombios_dp_get_dpcd only when error\n\nIn amdgpu_connector_hotplug(), we need to start DP link\ntraining only after we have received DPCD. The function\namdgpu_atombios_dp_get_dpcd() returns non-zero value only\nwhen an error condition is met, otherwise returns zero.\nSo in case the function encounters an error, we need to\nskip rest of the code and return from amdgpu_connector_hotplug()\nimmediately. Only when we are successfull in reading DPCD\npin, we should carry on with turning-on the monitor.\n\nSigned-off-by: Arindam Nath \u003carindam.nath@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "89913ea615754163ae9a066c9c1b95aa2a2f51b4", "tree": "a88cbf917deb754e8af133be080a0ac57d2efcb6", "parents": [ "370afa7ac5dc6c65cebeaee841ae700eb2387a55" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Mon Feb 29 16:11:07 2016 -0500" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:24 2016 -0500" }, "message": "drm/amdgpu/cz: remove commented out call to enable vce pg\n\nThis code path is not currently enabled now that we properly\nrespect the vce pg flags, so uncomment the actual pg calls\nso the code is as it should be we are eventually able to\nenable vce pg.\n\nReviewed-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Rex Zhu \u003cRex.Zhu@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "370afa7ac5dc6c65cebeaee841ae700eb2387a55", "tree": "bf5cc403b7ccc008b6d755356c074ea93d8b98ca", "parents": [ "b3dae7828399ef316e3fabf7e82c6415cb03a02e" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Mon Feb 29 15:29:48 2016 -0500" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:24 2016 -0500" }, "message": "drm/amdgpu/powerplay/cz: enable/disable vce dpm independent of vce pg\n\nIf we don\u0027t disable it when vce is not in use, we use extra power\nif vce pg is disabled.\n\nReviewed-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Rex Zhu \u003cRex.Zhu@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "b3dae7828399ef316e3fabf7e82c6415cb03a02e", "tree": "984ec134f860d91df86ee372c00878d068368c01", "parents": [ "9cac537332f5502c103415b25609548c276a09f8" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Thu Feb 25 11:24:52 2016 -0500" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:23 2016 -0500" }, "message": "drm/amdgpu/cz: enable/disable vce dpm even if vce pg is disabled\n\nI missed this when cleaning up the vce pg handling.\n\nReviewed-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Rex Zhu \u003cRex.Zhu@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "9cac537332f5502c103415b25609548c276a09f8", "tree": "63224379a79bf92b55da2b082b36010bd0363415", "parents": [ "feebe91aa9a9d99d9ec157612a614fadb79beb99" ], "author": { "name": "Chunming Zhou", "email": "David1.Zhou@amd.com", "time": "Mon Feb 29 14:12:38 2016 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:23 2016 -0500" }, "message": "drm/amdgpu/gfx8: specify which engine to wait before vm flush\n\nSelect between me and pfp properly.\n\nSigned-off-by: Chunming Zhou \u003cDavid1.Zhou@amd.com\u003e\nReviewed-by: Christian König \u003cchristian.koenig@amd.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "feebe91aa9a9d99d9ec157612a614fadb79beb99", "tree": "83ef320ca39d6b0360052418d6d77bc454a1d519", "parents": [ "4ea2efae0d117c5b5f44081bab941542d892e758" ], "author": { "name": "Christian König", "email": "christian.koenig@amd.com", "time": "Fri Feb 26 16:18:15 2016 +0100" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:22 2016 -0500" }, "message": "drm/amdgpu: apply gfx_v8 fixes to gfx_v7 as well\n\nWe never ported that back to CIK, so we could run into VM faults here.\n\nSigned-off-by: Christian König \u003cchristian.koenig@amd.com\u003e\nReviewed-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "4ea2efae0d117c5b5f44081bab941542d892e758", "tree": "f8e4fbd54c0d24788b26c86fbb9287438186ef8a", "parents": [ "dc26a2a2b3f037c2727f514b64e8a865721ec74b" ], "author": { "name": "Rex Zhu", "email": "Rex.Zhu@amd.com", "time": "Thu Feb 25 17:32:45 2016 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:22 2016 -0500" }, "message": "drm/amd/powerplay: send event to notify powerplay all modules are initialized.\n\nwith this event, powerplay can adjust current power state if needed.\n\nSigned-off-by: Rex Zhu \u003cRex.Zhu@amd.com\u003e\nReviewed-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "dc26a2a2b3f037c2727f514b64e8a865721ec74b", "tree": "98900d56d80c1962daad0c234a184b7e8f49ce15", "parents": [ "5e031d9fe8b0741f11d49667dfc3ebf5454121fd" ], "author": { "name": "Rex Zhu", "email": "Rex.Zhu@amd.com", "time": "Thu Feb 25 17:16:52 2016 +0800" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:21 2016 -0500" }, "message": "drm/amd/powerplay: export AMD_PP_EVENT_COMPLETE_INIT task to amdgpu.\n\nThis is needed to init the dynamic states without a display. To be\nused in the next commit.\n\nSigned-off-by: Rex Zhu \u003cRex.Zhu@amd.com\u003e\nReviewed-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\n" }, { "commit": "5e031d9fe8b0741f11d49667dfc3ebf5454121fd", "tree": "3431b8c4d5a802351f11a29ca0e64e5ffa5a73d4", "parents": [ "eda1d1cf8d18383f19cd2b752f786120efa4768f" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Feb 24 17:38:38 2016 -0500" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:21 2016 -0500" }, "message": "drm/radeon/pm: update current crtc info after setting the powerstate\n\nOn CI, we need to see if the number of crtcs changes to determine\nwhether or not we need to upload the mclk table again. In practice\nwe don\u0027t currently upload the mclk table again after the initial load.\nThe only reason you would would be to add new states, e.g., for\narbitrary mclk setting which is not currently supported.\n\nAcked-by: Christian König \u003cchristian.koenig@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "eda1d1cf8d18383f19cd2b752f786120efa4768f", "tree": "c6f6415c1160d2666782cf48c7c969db3328b37a", "parents": [ "6378076bcfdcd1d4f8d726d08d3fa044736873eb" ], "author": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Feb 24 17:18:25 2016 -0500" }, "committer": { "name": "Alex Deucher", "email": "alexander.deucher@amd.com", "time": "Wed Mar 02 11:01:20 2016 -0500" }, "message": "drm/amdgpu/pm: update current crtc info after setting the powerstate\n\nOn CI, we need to see if the number of crtcs changes to determine\nwhether or not we need to upload the mclk table again. In practice\nwe don\u0027t currently upload the mclk table again after the initial load.\nThe only reason you would would be to add new states, e.g., for\narbitrary mclk setting which is not currently supported.\n\nAcked-by: Jordan Lazare \u003cJordan.Lazare@amd.com\u003e\nAcked-by: Christian König \u003cchristian.koenig@amd.com\u003e\nSigned-off-by: Alex Deucher \u003calexander.deucher@amd.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "e1f33be9186363da7955bcb5f0b03e6685544c50", "tree": "351ee45f7f1e591fd755b38355a945e11547dd04", "parents": [ "4e94ebdd06d5dc72b7a40fc12fc496d601fb7bbc" ], "author": { "name": "Greg Kurz", "email": "gkurz@linux.vnet.ibm.com", "time": "Tue Feb 16 15:54:28 2016 +0100" }, "committer": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Wed Mar 02 17:01:49 2016 +0200" }, "message": "vhost: fix error path in vhost_init_used()\n\nWe don\u0027t want side effects. If something fails, we rollback vq-\u003eis_le to\nits previous value.\n\nSigned-off-by: Greg Kurz \u003cgkurz@linux.vnet.ibm.com\u003e\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\n" }, { "commit": "4e94ebdd06d5dc72b7a40fc12fc496d601fb7bbc", "tree": "113da9721a062241f4108c993d29bdb67740cd36", "parents": [ "fc77dbd34c5c99bce46d40a2491937c3bcbd10af" ], "author": { "name": "Ladi Prosek", "email": "lprosek@redhat.com", "time": "Mon Feb 01 19:36:31 2016 +0100" }, "committer": { "name": "Michael S. Tsirkin", "email": "mst@redhat.com", "time": "Wed Mar 02 17:01:49 2016 +0200" }, "message": "virtio-pci: read the right virtio_pci_notify_cap field\n\nLooks like a copy-paste bug. The value is used as an optimization and a\nwrong value probably isn\u0027t causing any serious damage. Found when\nporting this code to Windows.\n\nSigned-off-by: Ladi Prosek \u003clprosek@redhat.com\u003e\nSigned-off-by: Michael S. Tsirkin \u003cmst@redhat.com\u003e\n\n\n\n\n" } ], "next": "eda908967feecf7575a8ab74f1695b0445cf324e" }