)]}' { "log": [ { "commit": "d80e731ecab420ddcb79ee9d0ac427acbc187b4b", "tree": "e96a660b75b5bee8ae2c315878ec917b9c0da446", "parents": [ "855a85f704026d5fe7de94fb1b765fe03404507f" ], "author": { "name": "Oleg Nesterov", "email": "oleg@redhat.com", "time": "Fri Feb 24 20:07:11 2012 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Feb 24 11:42:50 2012 -0800" }, "message": "epoll: introduce POLLFREE to flush -\u003esignalfd_wqh before kfree()\n\nThis patch is intentionally incomplete to simplify the review.\nIt ignores ep_unregister_pollwait() which plays with the same wqh.\nSee the next change.\n\nepoll assumes that the EPOLL_CTL_ADD\u0027ed file controls everything\nf_op-\u003epoll() needs. In particular it assumes that the wait queue\ncan\u0027t go away until eventpoll_release(). This is not true in case\nof signalfd, the task which does EPOLL_CTL_ADD uses its -\u003esighand\nwhich is not connected to the file.\n\nThis patch adds the special event, POLLFREE, currently only for\nepoll. It expects that init_poll_funcptr()\u0027ed hook should do the\nnecessary cleanup. Perhaps it should be defined as EPOLLFREE in\neventpoll.\n\n__cleanup_sighand() is changed to do wake_up_poll(POLLFREE) if\n-\u003esignalfd_wqh is not empty, we add the new signalfd_cleanup()\nhelper.\n\nep_poll_callback(POLLFREE) simply does list_del_init(task_list).\nThis make this poll entry inconsistent, but we don\u0027t care. If you\nshare epoll fd which contains our sigfd with another process you\nshould blame yourself. signalfd is \"really special\". I simply do\nnot know how we can define the \"right\" semantics if it used with\nepoll.\n\nThe main problem is, epoll calls signalfd_poll() once to establish\nthe connection with the wait queue, after that signalfd_poll(NULL)\nreturns the different/inconsistent results depending on who does\nEPOLL_CTL_MOD/signalfd_read/etc. IOW: apart from sigmask, signalfd\nhas nothing to do with the file, it works with the current thread.\n\nIn short: this patch is the hack which tries to fix the symptoms.\nIt also assumes that nobody can take tasklist_lock under epoll\nlocks, this seems to be true.\n\nNote:\n\n\t- we do not have wake_up_all_poll() but wake_up_poll()\n\t is fine, poll/epoll doesn\u0027t use WQ_FLAG_EXCLUSIVE.\n\n\t- signalfd_cleanup() uses POLLHUP along with POLLFREE,\n\t we need a couple of simple changes in eventpoll.c to\n\t make sure it can\u0027t be \"lost\".\n\nReported-by: Maxime Bizon \u003cmbizon@freebox.fr\u003e\nCc: \u003cstable@kernel.org\u003e\nSigned-off-by: Oleg Nesterov \u003coleg@redhat.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "b8aeec34175fc8fe8b0d40efea4846dfc1ba663e", "tree": "8111b66ee029cad8950c58bd7d0e5df38a69a321", "parents": [ "6b0cd00bc396daf5c2dcf17a8d82055335341f46" ], "author": { "name": "Hidetoshi Seto", "email": "seto.hidetoshi@jp.fujitsu.com", "time": "Thu Oct 07 15:31:31 2010 +0900" }, "committer": { "name": "Andi Kleen", "email": "ak@linux.intel.com", "time": "Fri Oct 08 09:32:15 2010 +0200" }, "message": "HWPOISON/signalfd: add support for addr_lsb\n\nSimilar change as to signal delivery: copy out the si_addr_lsb field\nto user space in signalfd\n\nSigned-off-by: Hidetoshi Seto \u003cseto.hidetoshi@jp.fujitsu.com\u003e\nSigned-off-by: Andi Kleen \u003cak@linux.intel.com\u003e\n" }, { "commit": "a788fd53aec9a439f6b8bf57888c30aea1176e1b", "tree": "013149157888cb1cce5f02efa2d7def1f609e1ef", "parents": [ "68622c61dc7971382f5d69cd5d881e618ea30414" ], "author": { "name": "Jaswinder Singh Rajput", "email": "jaswinderrajput@gmail.com", "time": "Fri Jan 30 22:14:02 2009 +0530" }, "committer": { "name": "Jaswinder Singh Rajput", "email": "jaswinderrajput@gmail.com", "time": "Sat Jan 31 00:06:33 2009 +0530" }, "message": "headers_check fix: linux/signalfd.h\n\nfix the following \u0027make headers_check\u0027 warning:\n\n usr/include/linux/signalfd.h:19: found __[us]{8,16,32,64} type without #include \u003clinux/types.h\u003e\n\nSigned-off-by: Jaswinder Singh Rajput \u003cjaswinderrajput@gmail.com\u003e\n" }, { "commit": "5fb5e04926a54bc1c22bba7ca166840f4476196f", "tree": "999cf1d58e85ca801a03baeba114b9e8d2734249", "parents": [ "77d2720059618b9b6e827a8b73831eb6c6fad63c" ], "author": { "name": "Ulrich Drepper", "email": "drepper@redhat.com", "time": "Wed Jul 23 21:29:37 2008 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jul 24 10:47:29 2008 -0700" }, "message": "flag parameters: NONBLOCK in signalfd\n\nThis patch adds support for the SFD_NONBLOCK flag to signalfd4. The\nadditional changes needed are minimal.\n\nThe following test must be adjusted for architectures other than x86 and\nx86-64 and in case the syscall numbers changed.\n\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n#include \u003cfcntl.h\u003e\n#include \u003csignal.h\u003e\n#include \u003cstdio.h\u003e\n#include \u003cunistd.h\u003e\n#include \u003csys/syscall.h\u003e\n\n#ifndef __NR_signalfd4\n# ifdef __x86_64__\n# define __NR_signalfd4 289\n# elif defined __i386__\n# define __NR_signalfd4 327\n# else\n# error \"need __NR_signalfd4\"\n# endif\n#endif\n\n#define SFD_NONBLOCK O_NONBLOCK\n\nint\nmain (void)\n{\n sigset_t ss;\n sigemptyset (\u0026ss);\n sigaddset (\u0026ss, SIGUSR1);\n int fd \u003d syscall (__NR_signalfd4, -1, \u0026ss, 8, 0);\n if (fd \u003d\u003d -1)\n {\n puts (\"signalfd4(0) failed\");\n return 1;\n }\n int fl \u003d fcntl (fd, F_GETFL);\n if (fl \u003d\u003d -1)\n {\n puts (\"fcntl failed\");\n return 1;\n }\n if (fl \u0026 O_NONBLOCK)\n {\n puts (\"signalfd4(0) set non-blocking mode\");\n return 1;\n }\n close (fd);\n\n fd \u003d syscall (__NR_signalfd4, -1, \u0026ss, 8, SFD_NONBLOCK);\n if (fd \u003d\u003d -1)\n {\n puts (\"signalfd4(SFD_NONBLOCK) failed\");\n return 1;\n }\n fl \u003d fcntl (fd, F_GETFL);\n if (fl \u003d\u003d -1)\n {\n puts (\"fcntl failed\");\n return 1;\n }\n if ((fl \u0026 O_NONBLOCK) \u003d\u003d 0)\n {\n puts (\"signalfd4(SFD_NONBLOCK) does not set non-blocking mode\");\n return 1;\n }\n close (fd);\n\n puts (\"OK\");\n\n return 0;\n}\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\nSigned-off-by: Ulrich Drepper \u003cdrepper@redhat.com\u003e\nAcked-by: Davide Libenzi \u003cdavidel@xmailserver.org\u003e\nCc: Michael Kerrisk \u003cmtk.manpages@googlemail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "9deb27baedb79759c3ab9435a7d8b841842d56e9", "tree": "1c88393ba30db851ca0bb93c4e656d4e5dbb22b9", "parents": [ "7d9dbca34240ebb6ff88d8a29c6c7bffd098f0c1" ], "author": { "name": "Ulrich Drepper", "email": "drepper@redhat.com", "time": "Wed Jul 23 21:29:24 2008 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jul 24 10:47:27 2008 -0700" }, "message": "flag parameters: signalfd\n\nThis patch adds the new signalfd4 syscall. It extends the old signalfd\nsyscall by one parameter which is meant to hold a flag value. In this\npatch the only flag support is SFD_CLOEXEC which causes the close-on-exec\nflag for the returned file descriptor to be set.\n\nA new name SFD_CLOEXEC is introduced which in this implementation must\nhave the same value as O_CLOEXEC.\n\nThe following test must be adjusted for architectures other than x86 and\nx86-64 and in case the syscall numbers changed.\n\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n#include \u003cfcntl.h\u003e\n#include \u003csignal.h\u003e\n#include \u003cstdio.h\u003e\n#include \u003cunistd.h\u003e\n#include \u003csys/syscall.h\u003e\n\n#ifndef __NR_signalfd4\n# ifdef __x86_64__\n# define __NR_signalfd4 289\n# elif defined __i386__\n# define __NR_signalfd4 327\n# else\n# error \"need __NR_signalfd4\"\n# endif\n#endif\n\n#define SFD_CLOEXEC O_CLOEXEC\n\nint\nmain (void)\n{\n sigset_t ss;\n sigemptyset (\u0026ss);\n sigaddset (\u0026ss, SIGUSR1);\n int fd \u003d syscall (__NR_signalfd4, -1, \u0026ss, 8, 0);\n if (fd \u003d\u003d -1)\n {\n puts (\"signalfd4(0) failed\");\n return 1;\n }\n int coe \u003d fcntl (fd, F_GETFD);\n if (coe \u003d\u003d -1)\n {\n puts (\"fcntl failed\");\n return 1;\n }\n if (coe \u0026 FD_CLOEXEC)\n {\n puts (\"signalfd4(0) set close-on-exec flag\");\n return 1;\n }\n close (fd);\n\n fd \u003d syscall (__NR_signalfd4, -1, \u0026ss, 8, SFD_CLOEXEC);\n if (fd \u003d\u003d -1)\n {\n puts (\"signalfd4(SFD_CLOEXEC) failed\");\n return 1;\n }\n coe \u003d fcntl (fd, F_GETFD);\n if (coe \u003d\u003d -1)\n {\n puts (\"fcntl failed\");\n return 1;\n }\n if ((coe \u0026 FD_CLOEXEC) \u003d\u003d 0)\n {\n puts (\"signalfd4(SFD_CLOEXEC) does not set close-on-exec flag\");\n return 1;\n }\n close (fd);\n\n puts (\"OK\");\n\n return 0;\n}\n~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~\n\n[akpm@linux-foundation.org: add sys_ni stub]\nSigned-off-by: Ulrich Drepper \u003cdrepper@redhat.com\u003e\nAcked-by: Davide Libenzi \u003cdavidel@xmailserver.org\u003e\nCc: Michael Kerrisk \u003cmtk.manpages@googlemail.com\u003e\nCc: \u003clinux-arch@vger.kernel.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "14e4a0f2bb242f8008bc70b55fa834292c6a62af", "tree": "325c507c57e4cb25cc2659869297558ffbb0a2e6", "parents": [ "96532babc3e2ec4e0fce891d64319f183f043855" ], "author": { "name": "Robert P. J. Day", "email": "rpjday@crashcourse.ca", "time": "Sun Feb 03 15:12:15 2008 +0200" }, "committer": { "name": "Adrian Bunk", "email": "bunk@kernel.org", "time": "Sun Feb 03 15:12:15 2008 +0200" }, "message": "Fix a small number of \"memeber\" typoes.\n\nSigned-off-by: Robert P. J. Day \u003crpjday@crashcourse.ca\u003e\nSigned-off-by: Adrian Bunk \u003cbunk@kernel.org\u003e\n" }, { "commit": "96358de6bc2ab0bc5e8b44a8f3be4c9bf4a14e4f", "tree": "62f7ed295e7da3eb0706ca8d2b7454817bed6400", "parents": [ "059590f495f9c6e89cb018b9e612c3eec2336109" ], "author": { "name": "Davide Libenzi", "email": "davidel@xmailserver.org", "time": "Tue Oct 16 23:30:23 2007 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@woody.linux-foundation.org", "time": "Wed Oct 17 08:43:01 2007 -0700" }, "message": "rename signalfd_siginfo fields\n\nFor Michael Kerrisk request, the following patch renames signalfd_siginfo\nfields in order to keep them consistent with the siginfo_t ones.\n\nSigned-off-by: Davide Libenzi \u003cdavidel@xmailserver.org\u003e\nCc: Michael Kerrisk \u003cmtk-manpages@gmx.net\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "b8fceee17a310f189188599a8fa5e9beaff57eb0", "tree": "21308319be2579059a4d4d7db680a73334659f82", "parents": [ "9db619e66503494e41159de3c76fafabe80d016b" ], "author": { "name": "Davide Libenzi", "email": "davidel@xmailserver.org", "time": "Thu Sep 20 12:40:16 2007 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@woody.linux-foundation.org", "time": "Thu Sep 20 13:19:59 2007 -0700" }, "message": "signalfd simplification\n\nThis simplifies signalfd code, by avoiding it to remain attached to the\nsighand during its lifetime.\n\nIn this way, the signalfd remain attached to the sighand only during\npoll(2) (and select and epoll) and read(2). This also allows to remove\nall the custom \"tsk \u003d\u003d current\" checks in kernel/signal.c, since\ndequeue_signal() will only be called by \"current\".\n\nI think this is also what Ben was suggesting time ago.\n\nThe external effect of this, is that a thread can extract only its own\nprivate signals and the group ones. I think this is an acceptable\nbehaviour, in that those are the signals the thread would be able to\nfetch w/out signalfd.\n\nSigned-off-by: Davide Libenzi \u003cdavidel@xmailserver.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "fba2afaaec790dc5ab4ae8827972f342211bbb86", "tree": "2694d4cd8c6b7d69a5569b92151d61a3d4af39b7", "parents": [ "5dc8bf8132d59c03fe2562bce165c2f03f021687" ], "author": { "name": "Davide Libenzi", "email": "davidel@xmailserver.org", "time": "Thu May 10 22:23:13 2007 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@woody.linux-foundation.org", "time": "Fri May 11 08:29:36 2007 -0700" }, "message": "signal/timer/event: signalfd core\n\nThis patch series implements the new signalfd() system call.\n\nI took part of the original Linus code (and you know how badly it can be\nbroken :), and I added even more breakage ;) Signals are fetched from the same\nsignal queue used by the process, so signalfd will compete with standard\nkernel delivery in dequeue_signal(). If you want to reliably fetch signals on\nthe signalfd file, you need to block them with sigprocmask(SIG_BLOCK). This\nseems to be working fine on my Dual Opteron machine. I made a quick test\nprogram for it:\n\nhttp://www.xmailserver.org/signafd-test.c\n\nThe signalfd() system call implements signal delivery into a file descriptor\nreceiver. The signalfd file descriptor if created with the following API:\n\nint signalfd(int ufd, const sigset_t *mask, size_t masksize);\n\nThe \"ufd\" parameter allows to change an existing signalfd sigmask, w/out going\nto close/create cycle (Linus idea). Use \"ufd\" \u003d\u003d -1 if you want a brand new\nsignalfd file.\n\nThe \"mask\" allows to specify the signal mask of signals that we are interested\nin. The \"masksize\" parameter is the size of \"mask\".\n\nThe signalfd fd supports the poll(2) and read(2) system calls. The poll(2)\nwill return POLLIN when signals are available to be dequeued. As a direct\nconsequence of supporting the Linux poll subsystem, the signalfd fd can use\nused together with epoll(2) too.\n\nThe read(2) system call will return a \"struct signalfd_siginfo\" structure in\nthe userspace supplied buffer. The return value is the number of bytes copied\nin the supplied buffer, or -1 in case of error. The read(2) call can also\nreturn 0, in case the sighand structure to which the signalfd was attached,\nhas been orphaned. The O_NONBLOCK flag is also supported, and read(2) will\nreturn -EAGAIN in case no signal is available.\n\nIf the size of the buffer passed to read(2) is lower than sizeof(struct\nsignalfd_siginfo), -EINVAL is returned. A read from the signalfd can also\nreturn -ERESTARTSYS in case a signal hits the process. The format of the\nstruct signalfd_siginfo is, and the valid fields depends of the (-\u003ecode \u0026\n__SI_MASK) value, in the same way a struct siginfo would:\n\nstruct signalfd_siginfo {\n\t__u32 signo;\t/* si_signo */\n\t__s32 err;\t/* si_errno */\n\t__s32 code;\t/* si_code */\n\t__u32 pid;\t/* si_pid */\n\t__u32 uid;\t/* si_uid */\n\t__s32 fd;\t/* si_fd */\n\t__u32 tid;\t/* si_fd */\n\t__u32 band;\t/* si_band */\n\t__u32 overrun;\t/* si_overrun */\n\t__u32 trapno;\t/* si_trapno */\n\t__s32 status;\t/* si_status */\n\t__s32 svint;\t/* si_int */\n\t__u64 svptr;\t/* si_ptr */\n\t__u64 utime;\t/* si_utime */\n\t__u64 stime;\t/* si_stime */\n\t__u64 addr;\t/* si_addr */\n};\n\n[akpm@linux-foundation.org: fix signalfd_copyinfo() on i386]\nSigned-off-by: Davide Libenzi \u003cdavidel@xmailserver.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" } ] }