#
b5b023b9 |
| 07-Mar-2017 |
Hiren Panchasara <hiren@FreeBSD.org> |
We've found a recurring problem where some userland process would be stuck spinning at 100% cpu around sbcut_internal(). Inside sbflush_internal(), sb_ccc reached to about 4GB and before passing it t
We've found a recurring problem where some userland process would be stuck spinning at 100% cpu around sbcut_internal(). Inside sbflush_internal(), sb_ccc reached to about 4GB and before passing it to sbcut_internal(), we type-cast it from uint to int making it -ve.
The root cause of sockbuf growing this large is unknown. Correct fix is also not clear but based on mailing list discussions, adding KASSERTs to panic instead of looping endlessly.
Reviewed by: glebius Sponsored by: Limelight Networks
show more ...
|
Revision tags: vendor/NetBSD/bmake/20170301, vendor/acpica/20170303, vendor/libarchive/3.3.1, vendor/dma/20170210, vendor/ck/20170228, vendor/llvm/llvm-release_40-r296509, vendor/acpica/20170224, vendor/lld/lld-release_40-r296202, vendor/lld/lld-release_40-r296509, vendor/clang/clang-release_40-r296202, vendor/clang/clang-release_40-r296509, vendor/llvm/llvm-release_40-r296202, vendor/clang/clang-release_40-r296002, vendor/llvm/llvm-release_40-r296002, vendor/clang/clang-release_40-r295910, vendor/llvm/llvm-release_40-r295910, vendor/mandoc/1.4.1, vendor/openpam/RADULA, vendor/mandoc/1.4.1rc2, vendor/libucl/20170219, vendor/libc++/libc++-release_40-r295380, vendor/libc++/libc++-release_40-r295910, vendor/libc++/libc++-release_40-r296002, vendor/libc++/libc++-release_40-r296202, vendor/libc++/libc++-release_40-r296509, vendor/compiler-rt/compiler-rt-release_40-r295380, vendor/compiler-rt/compiler-rt-release_40-r295910, vendor/compiler-rt/compiler-rt-release_40-r296002, vendor/compiler-rt/compiler-rt-release_40-r296202, vendor/compiler-rt/compiler-rt-release_40-r296509, vendor/compiler-rt/compiler-rt-release_400-r297347, vendor/clang/clang-release_40-r295380, vendor/llvm/llvm-release_40-r295380, vendor/libpcap/1.8.1, vendor/lld/lld-release_40-r294803, vendor/lld/lld-release_40-r295380, vendor/lld/lld-release_40-r295910, vendor/lld/lld-release_40-r296002, vendor/libc++/libc++-release_40-r294803, vendor/clang/clang-release_40-r294803, vendor/llvm/llvm-release_40-r294803, vendor/lld/lld-release_40-r294123, vendor/libc++/libc++-release_40-r294123, vendor/compiler-rt/compiler-rt-release_40-r294123, vendor/compiler-rt/compiler-rt-release_40-r294803, vendor/clang/clang-release_40-r294123, vendor/llvm/llvm-release_40-r294123, vendor/NetBSD/tests/02.04.2017_10.12, vendor/unbound/1.6.0, vendor/ldns/1.7.0, vendor/byacc/20170201, vendor/lld/lld-release_40-r293807, vendor/libc++/libc++-release_40-r293807, vendor/clang/clang-release_40-r293807, vendor/llvm/llvm-release_40-r293807, vendor/tcpdump/4.9.0, vendor/openssh/7.4p1, vendor/openssh/7.3p1, vendor/NetBSD/libedit/2016-03-21, vendor/openresolv/3.9.0, vendor/lld/lld-release_40-r293443, vendor/libc++/libc++-release_40-r293443, vendor/compiler-rt/compiler-rt-release_40-r293443, vendor/compiler-rt/compiler-rt-release_40-r293807, vendor/clang/clang-release_40-r293443, vendor/llvm/llvm-release_40-r293443, vendor/openssl/1.0.2k, vendor/libc++/libc++-release_40-r292951, vendor/clang/clang-release_40-r292951, vendor/llvm/llvm-release_40-r292951, vendor/lld/lld-release_40-r292732, vendor/lld/lld-release_40-r292951, vendor/libc++/libc++-release_40-r292732, vendor/clang/clang-release_40-r292732, vendor/llvm/llvm-release_40-r292732, vendor/mandoc/20170121, zfs-0.7.0-rc3, vendor/acpica/20170119, vendor/NetBSD/tests/01.17.2017_21.34, vendor/zlib/1.2.11, vendor/lldb/lldb-release_40-r292009, vendor/lldb/lldb-release_40-r292732, vendor/lldb/lldb-release_40-r292951, vendor/lldb/lldb-release_40-r293443, vendor/lldb/lldb-release_40-r293807, vendor/lldb/lldb-release_40-r294123, vendor/lldb/lldb-release_40-r294803, vendor/lldb/lldb-release_40-r295380, vendor/lldb/lldb-release_40-r295910, vendor/lldb/lldb-release_40-r296002, vendor/lldb/lldb-release_40-r296202, vendor/lldb/lldb-release_40-r296509, vendor/lldb/lldb-release_400-r297347, vendor/lld/lld-release_40-r292009, vendor/libc++/libc++-release_40-r292009, vendor/compiler-rt/compiler-rt-release_40-r292009, vendor/compiler-rt/compiler-rt-release_40-r292732, vendor/compiler-rt/compiler-rt-release_40-r292951, vendor/clang/clang-release_40-r292009, vendor/llvm/llvm-release_40-r292009, vendor/NetBSD/tests/01.11.2017_23.20, vendor/lldb/lldb-trunk-r291476, vendor/lld/lld-trunk-r291476, vendor/libc++/libc++-trunk-r291476, vendor/compiler-rt/compiler-rt-trunk-r291476, vendor/clang/clang-trunk-r291476, vendor/llvm/llvm-trunk-r291476, vendor/lldb/lldb-trunk-r291274, vendor/lld/lld-trunk-r291274, vendor/libc++/libc++-trunk-r291274, vendor/compiler-rt/compiler-rt-trunk-r291274, vendor/clang/clang-trunk-r291274, vendor/llvm/llvm-trunk-r291274, vendor/xz/5.2.3, vendor/clang/clang-trunk-r291015, vendor/llvm/llvm-trunk-r291015, vendor/lldb/lldb-trunk-r291012, vendor/lldb/lldb-trunk-r291015, vendor/lld/lld-trunk-r291012, vendor/lld/lld-trunk-r291015, vendor/libc++/libc++-trunk-r291012, vendor/libc++/libc++-trunk-r291015, vendor/compiler-rt/compiler-rt-trunk-r291012, vendor/compiler-rt/compiler-rt-trunk-r291015, vendor/clang/clang-trunk-r291012, vendor/llvm/llvm-trunk-r291012, vendor/zlib/1.2.10, vendor/zlib/1.2.8-full, vendor/lldb/lldb-trunk-r290819, vendor/lld/lld-trunk-r290819, vendor/libc++/libc++-trunk-r290819, vendor/compiler-rt/compiler-rt-trunk-r290819, vendor/clang/clang-trunk-r290819, vendor/llvm/llvm-trunk-r290819, vendor/ck/20161230, vendor/heirloom-doctools/20161106, vendor/acpica/20161222, vendor/NetBSD/bmake/20161212, vendor/clang/clang-release_391-r289601, vendor/llvm/llvm-release_39-r288847, vendor/llvm/llvm-release_391-r289601, vendor/openbsm/1.2-ALPHA-5, vendor/byacc/20161202, vendor/tzdata/tzdata2016j, vendor/lld/lld-release_39-r288513, vendor/lld/lld-release_39-r288847, vendor/lld/lld-release_391-r289601, vendor/compiler-rt/compiler-rt-release_39-r288513, vendor/compiler-rt/compiler-rt-release_39-r288847, vendor/compiler-rt/compiler-rt-release_391-r289601, vendor/clang/clang-release_39-r288513, vendor/clang/clang-release_39-r288847, vendor/llvm/llvm-release_39-r288513, vendor/subversion/subversion-1.9.5, vendor/ck/20161128, vendor/lldb/lldb-release_39-r287912, vendor/lldb/lldb-release_39-r288513, vendor/lldb/lldb-release_39-r288847, vendor/lldb/lldb-release_391-r289601, vendor/lld/lld-release_39-r287912, vendor/libc++/libc++-release_39-r287912, vendor/libc++/libc++-release_39-r288513, vendor/libc++/libc++-release_39-r288847, vendor/libc++/libc++-release_391-r289601, vendor/compiler-rt/compiler-rt-release_39-r287912, vendor/clang/clang-release_39-r287912, vendor/llvm/llvm-release_39-r287912, vendor/ntp/4.2.8p9, vendor/acpica/20161117, vendor/file/5.29, vendor/tzdata/tzdata2016i, vendor/bind9/9.9.9-P4, zfs-0.7.0-rc2, vendor/heirloom-doctools/20161025, vendor/libarchive/3.2.2, vendor/heirloom-doctools/20161016, vendor/tzdata/tzdata2016h, vendor/tzdata/tzdata2016g, vendor/heirloom-doctools/20161006, vendor/byacc/20160606, vendor/libucl/20160812, vendor/dma/20160929, vendor/acpica/20160930, release/11.0.1, upstream/11.0.1, vendor/bind9/9.9.9-P3, vendor/unbound/1.5.10, vendor/openssl/1.0.2j, release/11.0.0, upstream/11.0.0, vendor/openssl/1.0.1u, vendor/openssl/1.0.2i, vendor/alpine-hal/2.7a, vendor/cortex-strings/linaro-eb80ac |
|
#
69a28758 |
| 15-Sep-2016 |
Ed Maste <emaste@FreeBSD.org> |
Renumber license clauses in sys/kern to avoid skipping #3
|
Revision tags: vendor/NetBSD/bmake/20160818, zfs-0.7.0-rc1, vendor/unbound/1.5.9, vendor/clang/clang-release_390-r280324, vendor/acpica/20160831, vendor/elftoolchain/elftoolchain-r3490, vendor/amd/6.2, vendor/llvm/llvm-release_39-r279689, vendor/llvm/llvm-release_390-r280324, vendor/lld/lld-release_39-r279689, vendor/lld/lld-release_390-r280324, vendor/lld/lldb-release_39-r279477, vendor/libc++/libc++-release_39-r279477, vendor/libc++/libc++-release_39-r279689, vendor/libc++/libc++-release_390-r280324, vendor/compiler-rt/compiler-rt-release_39-r279477, vendor/compiler-rt/compiler-rt-release_39-r279689, vendor/compiler-rt/compiler-rt-release_390-r280324, vendor/clang/clang-release_39-r279477, vendor/clang/clang-release_39-r279689, vendor/llvm/llvm-release_39-r279477, vendor/sqlite3/sqlite-3140100, vendor/dma/20160806, vendor/lldb/lldb-release_39-r278877, vendor/lldb/lldb-release_39-r279477, vendor/lldb/lldb-release_39-r279689, vendor/lldb/lldb-release_390-r280324, vendor/lld/lld-release_39-r278877, vendor/libc++/libc++-release_39-r278877, vendor/compiler-rt/compiler-rt-release_39-r278877, vendor/clang/clang-release_39-r278877, vendor/llvm/llvm-release_39-r278877, vendor/NetBSD/tests/08.11.2016_18.01, vendor/acpica/20160729, vendor/device-tree/devicetree-965f3718, vendor/libdivsufsort/0.0.2015.10.27, vendor/lldb/lldb-release_39-r276489, vendor/lld/lld-release_39-r276489, vendor/libc++/libc++-release_39-r276489, vendor/compiler-rt/compiler-rt-release_39-r276489, vendor/clang/clang-release_39-r276489, vendor/llvm/llvm-release_39-r276489, vendor/mandoc/1.13.4, vendor/openresolv/3.8.1, vendor/llvm-libunwind/libunwind-r272680, vendor/Juniper/libxo/0.6.3, vendor/expat/2.2.0, vendor/file/5.28, vendor/libarchive/3.2.1, vendor/ldns-host/hg-20160610-170001, vendor/ldns-host/hg-20160501-114105, vendor/NetBSD/libc-vis/20160608, vendor/NetBSD/bmake/20160606, vendor/NetBSD/bmake/20160604, vendor/libucl/20160604, vendor/ntp/4.2.8p8, vendor/NetBSD/blacklist/20160409, vendor/Juniper/libxo/0.6.2, vendor/acpica/20160527, vendor/skein/1.3, vendor/elftoolchain/elftoolchain-r3477, vendor/NetBSD/bmake/20160512, vendor/elftoolchain/elftoolchain-r3475, vendor/file/5.27, vendor/libarchive/3.2.0, vendor/libcxxrt/2016-03-29-516a65c109eb0a01e5e95fbef455eb3215135cef, vendor/openssl/1.0.1t, vendor/openssl/1.0.2h, vendor/subversion/subversion-1.9.4, vendor/ntp/4.2.8p7, vendor/acpica/20160422, vendor/netcat/5.9, vendor/file/5.26, vendor/openresolv/3.8.0, vendor/libucl/libucl-0.8.0, vendor/sqlite3/sqlite-3120100, vendor/Juniper/libxo/0.6.1, vendor/Juniper/libxo/0.6.0 |
|
#
b85f65af |
| 15-Apr-2016 |
Pedro F. Giffuni <pfg@FreeBSD.org> |
kern: for pointers replace 0 with NULL.
These are mostly cosmetical, no functional change.
Found with devel/coccinelle.
|
Revision tags: vendor/Juniper/libxo/0.4.7, vendor/Juniper/libxo/0.4.6, vendor/byacc/20160324, release/10.3.0, upstream/10.3.0, vendor/acpica/20160318, vendor/NetBSD/bmake/20160315, vendor/Juniper/libxo/0.4.5, vendor/NetBSD/bmake/20160307, vendor/openssh/7.2p2, vendor/openssh/7.2p1, vendor/bind9/9.9.8-P4, vendor/unbound/1.5.8, vendor/clang/clang-release_380-r262564, vendor/llvm/llvm-release_380-r262564 |
|
#
f3215338 |
| 01-Mar-2016 |
John Baldwin <jhb@FreeBSD.org> |
Refactor the AIO subsystem to permit file-type-specific handling and improve cancellation robustness.
Introduce a new file operation, fo_aio_queue, which is responsible for queueing and completing a
Refactor the AIO subsystem to permit file-type-specific handling and improve cancellation robustness.
Introduce a new file operation, fo_aio_queue, which is responsible for queueing and completing an asynchronous I/O request for a given file. The AIO subystem now exports library of routines to manipulate AIO requests as well as the ability to run a handler function in the "default" pool of AIO daemons to service a request.
A default implementation for file types which do not include an fo_aio_queue method queues requests to the "default" pool invoking the fo_read or fo_write methods as before.
The AIO subsystem permits file types to install a private "cancel" routine when a request is queued to permit safe dequeueing and cleanup of cancelled requests.
Sockets now use their own pool of AIO daemons and service per-socket requests in FIFO order. Socket requests will not block indefinitely permitting timely cancellation of all requests.
Due to the now-tight coupling of the AIO subsystem with file types, the AIO subsystem is now a standard part of all kernels. The VFS_AIO kernel option and aio.ko module are gone.
Many file types may block indefinitely in their fo_read or fo_write callbacks resulting in a hung AIO daemon. This can result in hung user processes (when processes attempt to cancel all outstanding requests during exit) or a hung system. To protect against this, AIO requests are only permitted for known "safe" files by default. AIO requests for all file types can be enabled by setting the new vfs.aio.enable_usafe sysctl to a non-zero value. The AIO tests have been updated to skip operations on unsafe file types if the sysctl is zero.
Currently, AIO requests on sockets and raw disks are considered safe and are enabled by default. aio_mlock() is also enabled by default.
Reviewed by: cem, jilles Discussed with: kib (earlier version) Sponsored by: Chelsio Communications Differential Revision: https://reviews.freebsd.org/D5289
show more ...
|
Revision tags: vendor/openssl/1.0.1s, vendor/openssl/1.0.2g, vendor/byacc/20150711, vendor/byacc/20141128, vendor/openresolv/3.7.3, vendor/NetBSD/libedit/2016-02-27, vendor/clang/clang-release_38-r261684, vendor/llvm/llvm-release_38-r261684, vendor/libc++/libc++-release_38-r261369, vendor/libc++/libc++-release_380-r262564, vendor/compiler-rt/compiler-rt-release_38-r261369, vendor/compiler-rt/compiler-rt-release_380-r262564, vendor/clang/clang-release_38-r261369, vendor/llvm/llvm-release_38-r261369, vendor/lldb/lldb-release_38-r260756, vendor/lldb/lldb-release_380-r262564, vendor/libc++/libc++-release_38-r260756, vendor/compiler-rt/compiler-rt-release_38-r260756, vendor/clang/clang-release_38-r260756, vendor/llvm/llvm-release_38-r260756, vendor/acpica/20160212, vendor/elftoolchain/elftoolchain-r3399, vendor/elftoolchain/elftoolchain-r3395, vendor/NetBSD/libedit/2016-01-30 |
|
#
8ec07310 |
| 01-Feb-2016 |
Gleb Smirnoff <glebius@FreeBSD.org> |
These files were getting sys/malloc.h and vm/uma.h with header pollution via sys/mbuf.h
|
Revision tags: vendor/openresolv/3.7.2, vendor/device-tree/ianc-afaecb70, vendor/openssl/1.0.1r, vendor/openssl/1.0.2f, vendor/lldb/lldb-release_38-r258968, vendor/lld/lld-release_38-r258968, vendor/lld/lld-release_380-r262564, vendor/libc++/libc++-release_38-r258968, vendor/compiler-rt/compiler-rt-release_38-r258968, vendor/clang/clang-release_38-r258968, vendor/llvm/llvm-release_38-r258968, vendor/alpine-hal/2.7, vendor/libc++/libc++-release_38-r258549, vendor/clang/clang-release_38-r258549, vendor/llvm/llvm-release_38-r258549, vendor/ntp/4.2.8p6, vendor/bind9/9.9.8-P3, vendor/openssh/7.1p2, vendor/lldb/lldb-release_38-r257836, vendor/libc++/libc++-release_38-r257836, vendor/clang/clang-release_38-r257836, vendor/llvm/llvm-release_38-r257836, vendor/mandoc/20160116, vendor/lldb/lldb-trunk-r257626, vendor/lld/lld-trunk-r257626, vendor/libc++/r257626, vendor/compiler-rt/compiler-rt-trunk-r257626, vendor/clang/clang-trunk-r257626, vendor/llvm/llvm-trunk-r257626, vendor/acpica/20160108 |
|
#
829fae90 |
| 08-Jan-2016 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Make it possible for sbappend() to preserve M_NOTREADY on mbufs, just like sbappendstream() does. Although, M_NOTREADY may appear only on SOCK_STREAM sockets, due to sendfile(2) supporting only the l
Make it possible for sbappend() to preserve M_NOTREADY on mbufs, just like sbappendstream() does. Although, M_NOTREADY may appear only on SOCK_STREAM sockets, due to sendfile(2) supporting only the latter, there is a corner case of AF_UNIX/SOCK_STREAM socket, that still uses records for the sake of control data, albeit being stream socket.
Provide private version of m_clrprotoflags(), which understands PRUS_NOTREADY, similar to m_demote().
show more ...
|
Revision tags: vendor/ntp/4.2.8p5, vendor/lldb/lldb-trunk-r256945, vendor/lld/lld-trunk-r256945, vendor/libc++/r256945, vendor/compiler-rt/compiler-rt-trunk-r256945, vendor/clang/clang-trunk-r256945, vendor/llvm/llvm-trunk-r256945, vendor/llvm-libunwind/libunwind-r256779, vendor/less/v481, vendor/lldb/lldb-trunk-r256633, vendor/llvm/llvm-trunk-r256633, vendor/lld/lld-trunk-r256633, vendor/libc++/r256633, vendor/compiler-rt/compiler-rt-trunk-r256633, vendor/clang/clang-trunk-r256633, vendor/clang/clang-release_371-r255217, vendor/llvm/llvm-release_371-r255217, vendor/NetBSD/bmake/20151220, vendor/acpica/20151218, vendor/bind9/9.9.8-P2, vendor/unbound/1.5.7, vendor/unbound/1.5.6, vendor/elftoolchain/elftoolchain-r3272, vendor/openbsm/1.2-ALPHA-4, vendor/NetBSD/bmake/20151201, vendor/openssl/0.9.8zh, vendor/openssl/1.0.1q, vendor/openssl/1.0.2e, vendor/acpica/20151124, vendor/xz/5.2.2, vendor/tzdata/tzdata2015g, vendor/libucl/20151027, vendor/openssl/1.0.2d, vendor/NetBSD/bmake/20151020, vendor/ntp/4.2.8p4, vendor/wpa/2.5, vendor/subversion/subversion-1.9.2, vendor/dma/0.10, vendor/elftoolchain/elftoolchain-r3250, vendor/unbound/1.5.5, vendor/bind9/9.9.8, vendor/acpica/20150930, vendor/netcat/5.8, vendor/llvm-libunwind/libunwind-r246528, vendor/file/5.25, vendor/unbound/1.5.4, zfs-0.6.5, vendor/libc++/libc++-release_370-r246257, vendor/compiler-rt/compiler-rt-release_370-r246257, vendor/lldb/lldb-release_370-r246257, vendor/clang/clang-release_370-r246257, vendor/llvm/llvm-release_370-r246257, vendor/lldb/lldb-trunk-r242221, vendor/file/5.24, vendor/openssh/7.1p1, vendor/openssh/7.0p1, vendor/acpica/20150818, vendor/pciids/pciids-20150815, vendor/tzdata/tzdata2015f, release/10.2.0, upstream/10.2.0, vendor/sqlite3/sqlite-3081101, vendor/subversion/subversion-1.8.14, vendor/serf/serf-1.3.8, vendor/apr/apr-1.5.2, vendor/clang/clang-trunk-r242221, vendor/llvm/llvm-trunk-r242221, vendor/bind9/9.9.7-P2, vendor/acpica/20150717, vendor/openssl/1.0.1p, vendor/sendmail/8.15.2, vendor/tcpdump/4.7.4, vendor/clang/clang-trunk-r241361, vendor/llvm/llvm-trunk-r241361, vendor/lldb/lldb-r241361, vendor/openssh/6.9p1, vendor/openssh/6.8p1, vendor/ntp/4.2.8p3, vendor/clang/clang-trunk-r240225, vendor/llvm/llvm-trunk-r240225, vendor/acpica/20150619, vendor/libcxxrt/2015-06-18-e64e93fe5bba67a6d52cbe5a97f8770c054bfa65, vendor/acpica/20150616, vendor/tzdata/tzdata2015e, vendor/openssl/1.0.1o, vendor/openssl/0.9.8zg, vendor/openssl/1.0.1n, vendor/NetBSD/bmake/20150606, vendor/file/5.23 |
|
#
f6f6d240 |
| 10-Jun-2015 |
Mateusz Guzik <mjg@FreeBSD.org> |
Implement lockless resource limits.
Use the same scheme implemented to manage credentials.
Code needing to look at process's credentials (as opposed to thred's) is provided with *_proc variants of
Implement lockless resource limits.
Use the same scheme implemented to manage credentials.
Code needing to look at process's credentials (as opposed to thred's) is provided with *_proc variants of relevant functions.
Places which possibly had to take the proc lock anyway still use the proc pointer to access limits.
show more ...
|
Revision tags: vendor/clang/clang-trunk-r239412, vendor/llvm/llvm-trunk-r239412, vendor/clang/clang-trunk-r238337, vendor/llvm/llvm-trunk-r238337, vendor/elftoolchain/elftoolchain-r3223, vendor/clang/clang-3.6.x, vendor/clang/clang-release_361-r237755, vendor/llvm/llvm-3.6.x, vendor/llvm/llvm-release_361-r237755, vendor/acpica/20150515, vendor/tzdata/tzdata2015d, vendor/tzdata/tzdata2015c, vendor/elftoolchain/elftoolchain-r3197, vendor/ficl/4.1.0, vendor/ficl/3.0.3, vendor/NetBSD/bmake/20150505, vendor/netcat/5.7, vendor/openresolv/3.7.0, vendor/sqlite3/sqlite-3080900, vendor/unbound/1.5.3, vendor/unbound/1.5.2, vendor/NetBSD/bmake/20150418, vendor/wpa/2.4, vendor/acpica/20150410, vendor/nvi/2.1.3, vendor/ntp/4.2.8p2, vendor/acpica/20150408, zfs-0.6.4, vendor/xz/5.2.1, vendor/bind9/9.9.7, vendor/elftoolchain/elftoolchain-r3179, vendor/lld/lld-trunk-r233088, vendor/tzdata/tzdata2015b, vendor/lua/5.3.0, vendor/openssl/0.9.8zf, vendor/openssl/1.0.1m, vendor/mandoc/1.13.3, vendor/tzdata/tzdata2015a, vendor/libucl/20150302, vendor/mandoc/20150302, vendor/device-tree/ian-c8c1b3a7, vendor/clang/clang-release_360-r230434, vendor/llvm/llvm-release_360-r230434, vendor/compiler-rt/compiler-rt-r230183, vendor/clang/clang-release_360-r229772, vendor/llvm/llvm-release_360-r229772, vendor/elftoolchain/elftoolchain-r3163, vendor/clang/clang-release_360-r229040, vendor/llvm/llvm-release_360-r229040, vendor/compiler-rt/compiler-rt-r228651, vendor/lldb/lldb-r228549, vendor/lldb/lldb-r225923, vendor/xz/5.2.0, vendor/ntp/4.2.8p1, vendor/acpica/20150204, vendor/libcxxrt/2014-12-31-1cb607e89f6135bbc10f3d3b6fba1f983e258dcc, vendor/clang/clang-release_360-r227651, vendor/llvm/llvm-release_360-r227651, vendor/clang/clang-release_360-r226102, vendor/llvm/llvm-release_360-r226102, vendor/openssl/0.9.8ze, vendor/openssl/1.0.1l, vendor/clang/clang-release_351-r225668, vendor/llvm/llvm-release_351-r225668, vendor/NetBSD/libedit/2014-01-07, vendor/openssl/0.9.8zd, vendor/openssl/1.0.1k, vendor/libc++/r224926, vendor/compiler-rt/compiler-rt-r224034, vendor/tcpdump/4.6.2, vendor/libpcap/1.6.2, vendor/openssh/6.7p1, vendor/file/5.22, vendor/unbound/1.5.1, vendor/unbound/1.5.0, vendor/elftoolchain/elftoolchain-r3136, vendor/libcxxrt/2014-12-06-00bc29eb6513624824a6d7db2ebc768a4216a604, vendor/mandoc/1.13.2, vendor/elftoolchain/elftoolchain-r3130 |
|
#
53b680ca |
| 22-Dec-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
In sbappend*() family of functions clear M_PROTO flags of incoming mbufs. sbappendstream() already does this in m_demote().
PR: 196174 Sponsored by: Nginx, Inc.
|
Revision tags: vendor/ntp/4.2.8 |
|
#
e834a840 |
| 20-Dec-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Revert r274494, r274712, r275955 and provide extra comments explaining why there could appear a zero-sized mbufs in socket buffers.
A proper fix would be to divorce record socket buffers and stream
Revert r274494, r274712, r275955 and provide extra comments explaining why there could appear a zero-sized mbufs in socket buffers.
A proper fix would be to divorce record socket buffers and stream socket buffers, and divorce pru_send that accepts normal data from pru_send that accepts control data.
show more ...
|
#
b7413232 |
| 20-Dec-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Add to sbappendstream_locked() a check against NULL mbuf, like it is done in sbappend_locked() and sbappendrecord_locked().
This is a quick fix to the panic introduced by r274712.
A proper solution
Add to sbappendstream_locked() a check against NULL mbuf, like it is done in sbappend_locked() and sbappendrecord_locked().
This is a quick fix to the panic introduced by r274712.
A proper solution should be to make sosend_generic() avoid calling pru_send() with NULL mbuf for the protocols that do not understand control messages. Those protocols that understand control messages, should be able to receive NULL mbuf, if control is non-NULL.
show more ...
|
Revision tags: vendor/bind9/9.9.6-P1, vendor/sendmail/8.15.1, vendor/file/5.21, vendor/tnftp/20141104, vendor/tnftp/20141031, vendor/mandoc/20141201 |
|
#
651e4e6a |
| 30-Nov-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Merge from projects/sendfile: extend protocols API to support sending not ready data: o Add new flag to pru_send() flags - PRUS_NOTREADY. o Add new protocol method pru_ready().
Sponsored by: Nginx,
Merge from projects/sendfile: extend protocols API to support sending not ready data: o Add new flag to pru_send() flags - PRUS_NOTREADY. o Add new protocol method pru_ready().
Sponsored by: Nginx, Inc. Sponsored by: Netflix
show more ...
|
#
0f9d0a73 |
| 30-Nov-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Merge from projects/sendfile:
o Introduce a notion of "not ready" mbufs in socket buffers. These mbufs are now being populated by some I/O in background and are referenced outside. This forces fol
Merge from projects/sendfile:
o Introduce a notion of "not ready" mbufs in socket buffers. These mbufs are now being populated by some I/O in background and are referenced outside. This forces following implications: - An mbuf which is "not ready" can't be taken out of the buffer. - An mbuf that is behind a "not ready" in the queue neither. - If sockbet buffer is flushed, then "not ready" mbufs shouln't be freed.
o In struct sockbuf the sb_cc field is split into sb_ccc and sb_acc. The sb_ccc stands for ""claimed character count", or "committed character count". And the sb_acc is "available character count". Consumers of socket buffer API shouldn't already access them directly, but use sbused() and sbavail() respectively. o Not ready mbufs are marked with M_NOTREADY, and ready but blocked ones with M_BLOCKED. o New field sb_fnrdy points to the first not ready mbuf, to avoid linear search. o New function sbready() is provided to activate certain amount of mbufs in a socket buffer.
A special note on SCTP: SCTP has its own sockbufs. Unfortunately, FreeBSD stack doesn't yet allow protocol specific sockbufs. Thus, SCTP does some hacks to make itself compatible with FreeBSD: it manages sockbufs on its own, but keeps sb_cc updated to inform the stack of amount of data in them. The new notion of "not ready" data isn't supported by SCTP. Instead, only a mechanical substitute is done: s/sb_cc/sb_ccc/. A proper solution would be to take away struct sockbuf from struct socket and allow protocols to implement their own socket buffers, like SCTP already does. This was discussed with rrs@.
Sponsored by: Netflix Sponsored by: Nginx, Inc.
show more ...
|
#
57f43a45 |
| 30-Nov-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
- Move sbcheck() declaration under SOCKBUF_DEBUG. - Improve SOCKBUF_DEBUG macros. - Improve sbcheck().
Sponsored by: Netflix Sponsored by: Nginx, Inc.
|
#
8967b220 |
| 30-Nov-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Make sballoc() and sbfree() functions. Ideally, they could be marked as static, but unfortunately Infiniband (ab)uses them.
Sponsored by: Nginx, Inc.
|
Revision tags: vendor/libucl/20141129, vendor/lldb/lldb-r216948, vendor/clang/clang-release_350-r216957, vendor/llvm/llvm-release_350-r216957, vendor/sqlite3/sqlite-3080702, vendor/mandoc/1.13.1, vendor/tzdata/tzdata2014j |
|
#
8146bcfe |
| 14-Nov-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
- Use NULL to compare a pointer. - Use KASSERT() instead of panic. - Remove useless 'continue', no need to restart cycle here.
Sponsored by: Nginx, Inc.
|
#
f274e256 |
| 14-Nov-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
There should not be zero length mbufs in socket buffers. The code comes from r1451, and thus can't be explained. A patch with explicit panic() here survived all tests.
Tested by: pho Sponsored by:
There should not be zero length mbufs in socket buffers. The code comes from r1451, and thus can't be explained. A patch with explicit panic() here survived all tests.
Tested by: pho Sponsored by: Nginx, Inc.
show more ...
|
Revision tags: release/10.1.0, upstream/10.1.0, vendor/acpica/20141107, vendor/compiler-rt/compiler-rt-r197381, vendor/atf/atf-0.21, vendor/NetBSD/mtree/20141028, vendor/netcat/5.6, vendor/tzdata/tzdata2014i, vendor/xz/5.0.7, vendor/openssl/0.9.8zc, vendor/openssl/1.0.1j, vendor/byacc/20141006, vendor/byacc/20141005, vendor/NetBSD/tests/09.30.2014_20.45, vendor/acpica/20140926, vendor/apr-util/apr-util-1.5.4 |
|
#
9fd573c3 |
| 22-Sep-2014 |
Hans Petter Selasky <hselasky@FreeBSD.org> |
Improve transmit sending offload, TSO, algorithm in general.
The current TSO limitation feature only takes the total number of bytes in an mbuf chain into account and does not limit by the number of
Improve transmit sending offload, TSO, algorithm in general.
The current TSO limitation feature only takes the total number of bytes in an mbuf chain into account and does not limit by the number of mbufs in a chain. Some kinds of hardware is limited by two factors. One is the fragment length and the second is the fragment count. Both of these limits need to be taken into account when doing TSO. Else some kinds of hardware might have to drop completely valid mbuf chains because they cannot loaded into the given hardware's DMA engine. The new way of doing TSO limitation has been made backwards compatible as input from other FreeBSD developers and will use defaults for values not set.
Reviewed by: adrian, rmacklem Sponsored by: Mellanox Technologies MFC after: 1 week
show more ...
|
Revision tags: vendor/openpam/OUROUPARIA, vendor/acpica/20140828, vendor/pjdfstest/abf03c3a47745d4521b0e4aa141317553ca48f91, vendor/NetBSD/libc-vis/20140908, vendor/device-tree/ianc-b78b6b80, vendor/tzdata/tzdata2014f, vendor/sqlite3/sqlite-3080500, vendor/serf/serf-1.3.7, vendor/subversion/subversion-1.8.10, vendor/resolver/9.5.0, vendor/openssl/0.9.8zb, vendor/openssl/1.0.1i, vendor/serf/serf-1.3.6, vendor/libucl/20140718, vendor/libucl/20140716, vendor/byacc/20140715, release/9.3.0 |
|
#
2827952e |
| 08-Jul-2014 |
Xin LI <delphij@FreeBSD.org> |
Don't leave the padding between the msg header and the cmsg data, and the padding after the cmsg data un-initialized.
Submitted by: tuexen Security: CVE-2014-3952 Security: FreeBSD-SA-14:17.kmem
|
Revision tags: vendor/NetBSD/bmake/20140620, vendor/acpica/20140627, vendor/file/5.19, vendor/tzdata/tzdata2014e, zfs-0.6.3, vendor/openssl/0.9.8za, vendor/openssl/1.0.1h, vendor/apr/apr-1.5.1, vendor/subversion/subversion-1.8.9, vendor/serf/serf-1.3.5, vendor/byacc/20140422, vendor/libucl/20140514, vendor/sendmail/8.14.9, vendor/unbound/1.4.22, vendor/unbound/1.4.21, vendor/ldns/1.6.17, vendor/tzdata/tzdata2014c, vendor/clang/clang-release_34-r208032, vendor/llvm/llvm-release_34-r208032, vendor/acpica/20140424, vendor/byacc/20140409, vendor/libucl/0.4.0, vendor/netcat/5.5, vendor/openssl/1.0.1g, vendor/tzdata/tzdata2014b, vendor/acpica/20140325, vendor/libucl/20140321, vendor/openssh/6.6p1, vendor/tzdata/tzdata2014a, vendor/libucl/20140302 |
|
#
8de34a88 |
| 06-Mar-2014 |
Alan Somers <asomers@FreeBSD.org> |
Fix PR kern/185813 "SOCK_SEQPACKET AF_UNIX sockets with asymmetrical buffers drop packets". It was caused by a check for the space available in a sockbuf, but it was checking the wrong sockbuf.
sys
Fix PR kern/185813 "SOCK_SEQPACKET AF_UNIX sockets with asymmetrical buffers drop packets". It was caused by a check for the space available in a sockbuf, but it was checking the wrong sockbuf.
sys/sys/sockbuf.h sys/kern/uipc_sockbuf.c Add sbappendaddr_nospacecheck_locked(), which is just like sbappendaddr_locked but doesn't validate the receiving socket's space. Factor out common code into sbappendaddr_locked_internal(). We shouldn't simply make sbappendaddr_locked check the space and then call sbappendaddr_nospacecheck_locked, because that would cause the O(n) function m_length to be called twice.
sys/kern/uipc_usrreq.c Use sbappendaddr_nospacecheck_locked for SOCK_SEQPACKET sockets, because the receiving sockbuf's size limit is irrelevant.
tests/sys/kern/unix_seqpacket_test.c Now that 185813 is fixed, pipe_128k_8k fails intermittently due to 185812. Make it fail every time by adding a usleep after starting the writer thread and before starting the reader thread in test_pipe. That gives the writer time to fill up its send buffer. Also, clear the expected failure message due to 185813. It actually said "185812", but that was a typo.
PR: kern/185813 Reviewed by: silence from freebsd-net@ and rwatson@ MFC after: 3 weeks Sponsored by: Spectra Logic Corporation
show more ...
|
Revision tags: vendor/xz-embedded/6a8a2364434763a033781f6b2a605ace9a021013, vendor/ncurses/5.9-20140222, vendor/ncurses/5.9-20110404, vendor/ncurses/5.9-20110404_stripped, vendor/device-tree/ianc-efa963ec, vendor/lldb/lldb-r202189, vendor/bind9/9.9.5, vendor/bind9/9.8.7, vendor/libucl/20140222, vendor/serf/serf-1.3.4, vendor/dma/20140213, vendor/subversion/subversion-1.8.8, vendor/lldb/lldb-r201577, vendor/acpica/20140214, vendor/atf/atf-0.20, vendor/atf/atf-0.19, vendor/mandoc/1.12.3, vendor/openssh/6.5p1, vendor/libc++/r197960, vendor/dtc/dtc-6a15eb23, vendor/sendmail/8.14.8, vendor/NetBSD/bmake/20140101, vendor/openssl/1.0.1f |
|
#
761a9a1f |
| 17-Jan-2014 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Fix comment.
|
Revision tags: release/10.0.0, upstream/10.0.0, vendor/elftoolchain/elftoolchain-r2974, vendor/acpica/20140114, vendor/byacc/20140101, vendor/clang/clang-release_34-r197956, vendor/llvm/llvm-release_34-r197956, vendor/clang/clang-release_34-r197841, vendor/llvm/llvm-release_34-r197841, vendor/tzdata/tzdata2013i, vendor/acpica/20131218, vendor/nvi/2.1.2-c80f493b0382d3c, vendor/ntp/4.2.6p5, vendor/lldb/lldb-r196322, vendor/lldb/lldb-r196259, vendor/apr-util/apr-util-1.5.3, vendor/subversion/subversion-1.8.5, vendor/NetBSD/mtree/20131121, vendor/atf/atf-0.18, vendor/netcat/5.4, vendor/atf/atf-0.17, vendor/acpica/20131115, vendor/nvi/2.1.2-95773e17e2751, vendor/openssh/6.4p1, vendor/subversion/subversion-1.8.4, vendor/lldb/lldb-r194122, vendor/tzdata/tzdata2013h, vendor/byacc/20130925, vendor/acpica/20130927, vendor/NetBSD/mtree/20131016 |
|
#
1d2df300 |
| 09-Oct-2013 |
Gleb Smirnoff <glebius@FreeBSD.org> |
- Substitute sbdrop_internal() with sbcut_internal(). The latter doesn't free mbufs, but return chain of free mbufs to a caller. Caller can either reuse them or return to allocator in a batch man
- Substitute sbdrop_internal() with sbcut_internal(). The latter doesn't free mbufs, but return chain of free mbufs to a caller. Caller can either reuse them or return to allocator in a batch manner. - Implement sbdrop()/sbdrop_locked() as a wrapper around sbcut_internal(). - Expose sbcut_locked() for outside usage.
Sponsored by: Netflix Sponsored by: Nginx, Inc. Approved by: re (marius)
show more ...
|
Revision tags: vendor/dialog/1.2-20130923, vendor/tzdata/tzdata2013f, release/9.2.0, vendor/libcxxrt/2013-09-23-dafd555f547386c8c25f9afd07ea3693db13e52a, vendor/openssh/6.3p1, vendor/lldb/lldb-r188801, vendor/openpam/NUMMULARIA, vendor/NetBSD/bmake/20130904 |
|
#
7729cbf1 |
| 01-Sep-2013 |
Davide Italiano <davide@FreeBSD.org> |
Fix socket buffer timeouts precision using the new sbintime_t KPI instead of relying on the tvtohz() workaround. The latter has been introduced lately by jhb@ (r254699) in order to have a fix that ca
Fix socket buffer timeouts precision using the new sbintime_t KPI instead of relying on the tvtohz() workaround. The latter has been introduced lately by jhb@ (r254699) in order to have a fix that can be backported to STABLE.
Reported by: Vitja Makarov <vitja.makarov at gmail dot com> Reviewed by: jhb (earlier version)
show more ...
|
Revision tags: vendor/NetBSD/libexecinfo/20130829, vendor/ldns-host/hg-20120826-233833, vendor/acpica/20130823, zfs-0.6.2, vendor/NetBSD/libexecinfo/20130822, vendor/ipfilter/5.1.2, vendor/ipfilter-sys/5-1-2, vendor/bind9/9.9.3-P2, vendor/nvi/2.1.1-4334a8297f, vendor/serf/serf-1.3.0, vendor/NetBSD/bmake/20130730, vendor/bind9/9.8.5-P2, vendor/apr-util/apr-util-1.5.2, vendor/apr/apr-1.4.8, vendor/subversion/subversion-1.8.1, vendor/misc-GNU/patch/2.5.9, vendor/acpica/20130725, vendor/bind9/9.8.5-P1, vendor/libc++/r185801, vendor/libcxxrt/2013-07-08-c61efa043b14378efbd69c9a2686d44ed46ae179, vendor/v4l/2.6.34.14, vendor/hyperv/20130627, vendor/acpica/20130626, vendor/wpa/2.0 |
|
#
0963c8e4 |
| 19-Jun-2013 |
Lawrence Stewart <lstewart@FreeBSD.org> |
When a previous call to sbsndptr() leaves sb->sb_sndptroff at the start of an mbuf that was fully consumed by the previous call, the mbuf ptr returned by the current call ends up being the previous m
When a previous call to sbsndptr() leaves sb->sb_sndptroff at the start of an mbuf that was fully consumed by the previous call, the mbuf ptr returned by the current call ends up being the previous mbuf in the sb chain to the one that contains the data we want.
This does not cause any observable issues because the mbuf copy routines happily walk the mbuf chain to get to the data at the moff offset, which in this case means they effectively skip over the mbuf returned by sbsndptr().
We can't adjust sb->sb_sndptr during the previous call for this case because the next mbuf in the chain may not exist yet. We therefore need to detect the condition and make the adjustment during the current call.
Fix by detecting the special case of moff being at the start of the next mbuf in the chain and adjust the required accounting variables accordingly.
Reviewed by: andre MFC after: 2 weeks
show more ...
|
Revision tags: vendor/subversion/subversion-1.8.0, vendor/sqlite3/sqlite-3071700, vendor/subversion/subversion-1.8.0-rc3, vendor/serf/serf-1.2.1, vendor/apr-util/apr-util-1.4.1, vendor/apr/apr-1.4.6, vendor/dialog/1.2-20130523, vendor/clang/clang-release_33-r183502, vendor/llvm/llvm-release_33-r183502, vendor/NetBSD/bmake/20130604, release/8.4.0, vendor/byacc/20130304, vendor/tcpdump/4.4.0, vendor/libpcap/1.4.0, vendor/compiler-rt/compiler-rt-r182741, vendor/NetBSD/bmake/20130520, vendor/acpica/20130517, vendor/openssh/6.2p2, vendor/libregex/glibc-2.17, vendor/less/v458, vendor/tnftp/20130505, vendor/zlib/1.2.8, vendor/hyperv/20130502, vendor/flex/2.5.37, vendor/qcamain_open_hal/60390a9f9ac6a20db168fbbc01a4ad4e01c395ce, vendor/libc++/r180598, vendor/libcxxrt/2013-04-22-c812a07cd2f95c1403baf0bbe0366e7618d1d6d3, vendor/sendmail/8.14.7, vendor/tzdata/tzdata2013c, vendor/acpica/20130418, vendor/netcat/5.3, vendor/NetBSD/mtree/20130408, vendor/clang/clang-trunk-r178860, vendor/llvm/llvm-trunk-r178860, vendor/unbound/1.4.20, vendor/acpica/20130328, vendor/NetBSD/bmake/20130330 |
|
#
844cacd1 |
| 29-Mar-2013 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Once ng_ksocket(4) is fixed, re-apply r194662. See this revision for longer description.
Discussed with: andre, rwatson Sponsored by: Nginx, Inc.
|
Revision tags: vendor/bind9/9.8.4-P2, zfs-0.6.1, vendor/openssh/6.2p1, vendor/libarchive/3.1.2 |
|
#
c8b59ea7 |
| 15-Mar-2013 |
Gleb Smirnoff <glebius@FreeBSD.org> |
Use m_get() and m_getcl() instead of compat macros.
|