Home
last modified time | relevance | path

Searched full:overhead (Results 1 – 25 of 685) sorted by relevance

12345678910>>...28

/freebsd/libexec/rtld-elf/
H A Drtld_malloc.c66 * The overhead on a block is at least 4 bytes. When free, this space
71 union overhead { union
72 union overhead *ov_next; /* when free */
89 * (FIRST_BUCKET_SIZE << i). The overhead information precedes the data
95 static union overhead *nextf[NBUCKETS];
110 return (((caddr_t)cp - sizeof(union overhead))); in cp2op()
116 union overhead *op; in __crt_malloc()
174 union overhead ov1; in __crt_aligned_alloc_offset()
180 mem = __crt_malloc(size + align + offset + sizeof(union overhead)); in __crt_aligned_alloc_offset()
183 x = roundup2((uintptr_t)mem + sizeof(union overhead), align); in __crt_aligned_alloc_offset()
[all …]
/freebsd/contrib/tcsh/
H A Dtc.alloc.c97 * The overhead on a block is at least 4 bytes. When free, this space
109 union overhead { union
110 union overhead *ov_next; /* when free */
142 * smallest allocatable block is 8 bytes. The overhead information
146 static union overhead *nextf[NBUCKETS] IZERO_STRUCT;
155 static int findbucket (union overhead *, int);
180 union overhead *p; in malloc()
201 nbytes = MEMALIGN(MEMALIGN(sizeof(union overhead)) + nbytes + RSLOP); in malloc()
238 return ((memalign_t) (((caddr_t) p) + MEMALIGN(sizeof(union overhead)))); in malloc()
254 union overhead *op; in morecore()
[all …]
/freebsd/share/doc/papers/sysperf/
H A Da1.t53 * System call overhead benchmark.
78 * To calculate system overhead for a context
80 * with nsigs. Overhead is then estimated by
83 * overhead = t1 - 2 * t2;
291 * overhead (approximately). Process
294 * in calculating exec overhead.
345 * overhead (approximately). Process
349 * estimate the overhead for the exec.
H A D3.t78 The overhead of the routing and status updates can consume
80 Remote logins and shells incur more overhead
87 terminal handler overhead can roughly triple the load presented by one
90 System overhead
136 switch overhead. The \fIexec\fP tests use two different jobs to gauge
249 and context switch overhead did not change significantly
289 this factor to adjust the basic overhead of an \fIexec\fP system
369 The overhead of simply trapping and returning
391 this can result in excessive overhead.
403 is the basic overhead for clock processing.
[all …]
H A D4.t303 interrupt basis, avoiding the overhead
314 the overhead of checking the silos is incurred only
368 To reduce this overhead,
460 The overhead of packet routing has been decreased by changes in the routing
510 the overhead for system calls decreased by an average of 13%.
646 overhead.
688 This yielded another 10% cut in the basic overhead of delivering mail.
690 compiling it without debugging code reduced the overhead by another 20%.
713 as many as a dozen processes, easing system overhead and
722 we found that the overhead of the routing process was becoming
H A Dappendix.ms127 \fI*\fP System call overhead benchmark\&.
159 \fI*\fP To calculate system overhead for a context
161 \fI*\fP with nsigs\&. Overhead is then estimated by
164 \fI*\fP\h'|11n'overhead = t1 \*- 2 \fI*\fP t2;
413 \fI*\fP overhead (approximately)\&. Process
416 \fI*\fP in calculating exec overhead\&.
474 \fI*\fP overhead (approximately)\&. Process
478 \fI*\fP estimate the overhead for the exec\&.
H A D2.t174 The overhead for running a profiled system varies;
231 system operations. Operations such as system call overhead and
235 the overhead of these low level operations had not changed appreciably.
H A D7.t35 4.1BSD. This added overhead combined with problems with several
41 added system overhead by replacing old algorithms
/freebsd/sys/netgraph/
H A Dng_pipe.c99 u_int32_t overhead;
112 + priv->overhead ) * hinfo->run.fifo_queues * \
358 cfg->overhead = priv->overhead; in ngp_rcvmsg()
383 priv->overhead = 0; in ngp_rcvmsg()
389 priv->overhead = 8+4+12; /* Ethernet */ in ngp_rcvmsg()
391 priv->overhead = 10; /* HDLC */ in ngp_rcvmsg()
394 if (cfg->overhead == -1) in ngp_rcvmsg()
395 priv->overhead = 0; in ngp_rcvmsg()
396 else if (cfg->overhead > in ngp_rcvmsg()
98 u_int32_t overhead; global() member
[all...]
H A Dng_pipe.h147 u_int32_t overhead; member
157 { "overhead", &ng_parse_uint32_type }, \
/freebsd/crypto/openssl/ssl/record/
H A Dtls_pad.c78 const size_t overhead = 1 /* padding length byte */ + mac_size; in ssl3_cbc_remove_padding_and_mac() local
83 if (overhead > *reclen) in ssl3_cbc_remove_padding_and_mac()
87 good = constant_time_ge_s(*reclen, padding_length + overhead); in ssl3_cbc_remove_padding_and_mac()
124 size_t overhead = ((block_size == 1) ? 0 : 1) /* padding length byte */ in tls1_cbc_remove_padding_and_mac() local
131 if (overhead > *reclen) in tls1_cbc_remove_padding_and_mac()
144 good = constant_time_ge_s(*reclen, overhead + padding_length); in tls1_cbc_remove_padding_and_mac()
/freebsd/crypto/openssl/include/openssl/
H A Dssl3.h178 /* Maximum compression overhead: defined by SSL/TLS standards */
182 * The standards give a maximum encryption overhead of 1024 bytes. In
183 * practice the value is lower than this. The overhead is the maximum number
191 * send overhead is smaller.
197 /* If compression isn't used don't include the compression overhead */
/freebsd/contrib/capsicum-test/
H A Doverhead.cc37 FORK_TEST(Overhead, GetTid) { in FORK_TEST() argument
40 FORK_TEST(Overhead, Seek) { in FORK_TEST() argument
/freebsd/contrib/lua/src/
H A Dlapi.h38 ** To reduce the overhead of returning from C functions, the presence of
41 ** with zero, one, or "all" wanted results have no overhead. Functions
/freebsd/share/doc/smm/05.fastfs/
H A D4.t77 the basic system call overhead was a negligible portion of
140 The overhead of allocating blocks in the new system is greater
141 than the overhead of allocating blocks in the old system,
235 overhead of allocating at each write,
/freebsd/usr.bin/mkimg/
H A Dvmdk.c61 uint64_t overhead; member
176 fprintf(stderr, "VMDK: overhead = %ju\n", in vmdk_write()
179 le64enc(&hdr.overhead, sec); in vmdk_write()
/freebsd/contrib/llvm-project/llvm/include/llvm/Analysis/
H A DIndirectCallVisitor.h49 // non-vtable instructions and reduce instrumentation overhead. in tryGetVTableInstruction()
54 // of symbols). So the performance overhead from non-vtable profiled in tryGetVTableInstruction()
/freebsd/share/man/man9/
H A Dmtx_pool.9102 No structural overhead;
108 No initialization or destruction overhead.
/freebsd/contrib/llvm-project/llvm/include/llvm/ExecutionEngine/Orc/
H A DDebugObjectManagerPlugin.h61 /// overhead for objects built with a release configuration.
65 /// mode, but it may cause significant overhead when adding many modules in
/freebsd/crypto/heimdal/appl/ftp/ftp/
H A Dsecurity.h54 int (*overhead)(void *, int, int); member
65 int (*overhead)(void *, int, int); member
/freebsd/share/man/man4/
H A Dng_pipe.4123 u_int32_t overhead; /* assumed L2 overhead in bytes */
/freebsd/share/doc/papers/malloc/
H A Dperformance.ms16 We will refer to this as ``overhead time''.
21 The overhead time is easy to measure, just do a lot of malloc/free calls
/freebsd/lib/libc/include/
H A Dreentrant.h36 * be used by non-threaded applications without unreasonable overhead.
55 * it should not add significant run time or code size overhead to non-
/freebsd/share/doc/papers/newvm/
H A D1.t151 Another effect of the current technology is that the latency and overhead
239 the overhead of reading, flushing, and possibly allocating a file
247 nor be willing to pay the overhead associated with them.
266 to avoid a memory to memory copy, the overhead of doing the system
/freebsd/sys/contrib/openzfs/include/sys/
H A Dbtree.h30 * tree can be used to store arbitrary sortable data types with low overhead
65 * them, and increased memory overhead. Increasing these values results in
66 * higher variance in operation time, and reduces memory overhead.

12345678910>>...28