/linux-master/arch/m68k/fpsp040/ |
H A D | ssinh.S | 13 | Accuracy and Monotonicity: The returned result is within 3 ulps in 15 | result is subsequently rounded to double precision. The 16 | result is provably monotonic in double precision. 25 | 2. (|X| <= 16380 log2) Sinh(X) is obtained by the formulae 45 | the appropriate sign. Huge is the largest finite number in
|
H A D | sasin.S | 13 | Accuracy and Monotonicity: The returned result is within 3 ulps in 15 | result is subsequently rounded to double precision. The 16 | result is provably monotonic in double precision.
|
H A D | satanh.S | 14 | Accuracy and Monotonicity: The returned result is within 3 ulps in 16 | result is subsequently rounded to double precision. The 17 | result is provably monotonic in double precision.
|
H A D | round.S | 35 | On return the value pointed to by a0 is correctly rounded, 36 | a0 is preserved and the g-r-s bits in d0 are cleared. 37 | The result is not typed - the tag field is invalid. The 38 | result is still in the internal extended format. 46 | If g=r=s=0 then result is exact and round is done, else set 130 | selected rounding precision. It is called by the round subroutine 152 andil #0x0000003f,%d2 |s bit is the or of all other 164 andil #0x000001ff,%d2 |s bit is th [all...] |
H A D | binstr.S | 15 | The 64-bit binary is assumed to have a decimal point before 16 | bit 63. The fraction is multiplied by 10 using a mul by 2 18 | msb form a decimal digit. This process is iterated until 21 | A1. Init d7 to 1. D7 is the byte digit counter, and if 1, the 22 | digit formed will be assumed the least significant. This is 38 | A6. Test d7. If zero, the digit formed is the ms digit. If non- 39 | zero, it is the ls digit. Put the digit in its place in the 40 | upper word of d0. If it is the ls digit, write the word
|
H A D | x_fline.S | 6 | First determine if the exception is one of the unimplemented 8 | Next, determine if the instruction is an fmovecr with a non-zero 35 | the equate is valid only after a 'link a6' has pushed one more 63 cmpib #0x17,%d1 |check if it is an FMOVECR encoding
|
/linux-master/arch/arm/mach-pxa/ |
H A D | sleep.S | 10 * This program is free software; you can redistribute it and/or 62 @ set SDCLKx divide-by-2 bits (this is part of a workaround for Errata 50) 120 @ check that the memory frequency is within limits 128 @ note, turbo is not preserved over sleep so there is no 160 @ external accesses after SDRAM is put in self-refresh mode
|
/linux-master/arch/arc/kernel/ |
H A D | entry.S | 66 ; This is the historic "kernel_execve" use-case, to return to init 67 ; user mode, in a round about way since that is always done from 68 ; a kernel thread which is executed via JL above but always returns 69 ; out whenever kernel_execve (now inline do_fork()) is involved 148 ; for traps, ERET is pre-commit so points to next-PC 264 ; Disable IRQs to ensures that chk for pending work itself is atomic 295 ; (ii) If signal is SIGTRAP/SIGSTOP, task is being traced thus 323 ; CONFIG_PREEMPTION: This is a must for preempt_schedule_irq() 324 ; !CONFIG_PREEMPTION: To ensure restore_regs is int [all...] |
/linux-master/arch/s390/boot/ |
H A D | head.S | 44 brctg %r1,.Lnoload # is valid 54 larl %r2,_end # ramdisk loc. is temp 66 clc 0(3,%r4),0(%r13) # if it is HDRx 69 clc 0(3,%r4),0(%r13) # if it is EOFx 107 stsch 0(%r5) # check if irq is pending 110 tm 31(%r5),0xff # bits is set in the schib 215 # this is called either by the ipl loader or directly by PSW restart 223 # This is a list of s390 kernel entry points. At address 0x1000f the number of 224 # valid entry points is stored. 226 # IMPORTANT: Do not change this table, it is s39 [all...] |
/linux-master/arch/alpha/lib/ |
H A D | callback_srm.S | 19 ldq $0,hwrpb # gp is set up by CALLBACK macro. 104 .long 7 # value is not 1 for link debugging 107 .long 7 # value is not 1 for link debugging
|
H A D | ev6-memset.S | 5 * This is an efficient (and relatively small) implementation of the C library 26 * WARNING: Maintaining this is going to be more work than the above version, 28 * is worth it. 47 * Serious stalling happens. The only way to mitigate this is to 49 * with other parts of the fall-through code. This is important, even 76 beq $3,aligned_b # U : target is 0mod8 79 * Target address is misaligned, and won't fit within a quadword 86 addq $18,$3,$18 # E : $18 is new count ($3 is negative) 88 subq $16,$3,$16 # E : $16 is ne [all...] |
/linux-master/arch/m68k/ifpsp060/src/ |
H A D | fpsp.S | 9 THE SOFTWARE is provided on an "AS IS" basis and without warranty. 24 so long as this entire notice is retained without alteration in any modified and/or 31 # This file is appended to the top of the 060FPSP package 605 # - The system stack is unchanged # 608 # - The system stack is unchanged # 609 # - The "exception present" flag in the fsave frame is cleared # 612 # On the 060, if an FP overflow is present as the result of any # 614 # exception is enabled or disabled in the FPCR. For the disabled case, # 616 # default result should be for the operation. This default result is # 620 # If the exception is enable [all...] |
/linux-master/arch/x86/kernel/ |
H A D | verify_cpu.S | 13 * This is a common code for verification whether CPU supports 14 * long mode and SSE or not. It is not called directly instead this 15 * file is included at various places and compiled in that context. 16 * This file is expected to run in 32bit code. Currently: 52 movl $0x0,%eax # See if cpuid 1 is implemented 64 mov $1,%di # cpu is from AMD 106 movl $0x80000000,%eax # See if extended cpuid is implemented
|
/linux-master/arch/arm/probes/kprobes/ |
H A D | core.c | 56 int is; local 95 for (is = 0; is < MAX_INSN_SIZE; ++is) 96 p->ainsn.insn[is] = tmp_insn[is]; 149 * The actual disarming is done here on each CPU and synchronized using 150 * stop_machine. This synchronization is necessary on SMP to avoid removing 151 * a probe between the moment the 'Undefined Instruction' exception is raised 153 * memory. It is als [all...] |
/linux-master/arch/arm/common/ |
H A D | mcpm_head.S | 70 * MMU is off so we need to get to various variables in a 90 @ Signal that this CPU is coming UP: 97 @ state, because there is at least one active CPU (this CPU). 120 @ Signal that the cluster is being brought up: 145 @ power_up_setup is now responsible for setting up the cluster: 178 @ If a platform-specific CPU setup hook is needed, it is
|
/linux-master/arch/x86/lib/ |
H A D | checksum_32.S | 4 * operating system. INET is implemented using the BSD Socket 44 * is aligned on either a 2-byte or 4-byte boundary. We get at 45 * least a twofold speedup on 486 and Pentium if it is 4-byte aligned. 46 * Fortunately, it is easy to convert 2-byte alignment to 4-byte 56 jz 2f # Jump if alignment is ok. 58 jz 10f # Jump if alignment is boundary of 2 bytes. 60 # buf is odd 166 # buf is odd 238 andl -128(%esi),%ebx # esi is 4-aligned so should be ok 283 jz 2f # Jump if alignment is o [all...] |
/linux-master/arch/arm/mm/ |
H A D | l2c-l2x0-resume.S | 35 @ and can be written whether or not the L2 cache is enabled 43 @ Don't setup the L2 cache if it is already enabled
|
/linux-master/arch/x86/realmode/rm/ |
H A D | trampoline_32.S | 8 * This is only used for booting secondary CPUs in SMP machine 13 * is a mystery. 17 * On entry to trampoline_start, the processor is in real mode 19 * and IP is zero. Thus, we load CS to the physical segment 47 * operand size is 16bit. Use lgdtl instead to force operand size 51 lgdtl tr_gdt # load gdt with whatever is appropriate
|
/linux-master/arch/csky/kernel/ |
H A D | atomic.S | 41 sync.is
|
/linux-master/lib/ |
H A D | sg_pool.c | 19 #error SG_CHUNK_SIZE is too small (must be 32 or greater) 31 #error SG_CHUNK_SIZE is too large (256 MAX) 106 * Allocate and chain SGLs in an sg table. If @nents@ is larger than 107 * @nents_first_chunk a chained sg table will be setup. @first_chunk is
|
/linux-master/scripts/kconfig/ |
H A D | merge_config.sh | 121 SED_CONFIG_EXP2="s/^# \(${CONFIG_PREFIX}[a-zA-Z0-9_]*\) is not set$/\1/p" 154 ${WARNOVERRIDE} Value of $CFG is redefined by fragment $ORIG_MERGE_FILE: 162 ${WARNOVERRIDE} Value of $CFG is redundant by fragment $ORIG_MERGE_FILE: 196 # allnoconfig: Fills in any missing symbols with # CONFIG_* is not set
|
/linux-master/arch/xtensa/lib/ |
H A D | usercopy.S | 15 * This file is subject to the terms and conditions of the GNU General 26 * The returned value is the number of bytes not copied. Implies zero 27 * is success. 29 * The general case algorithm is as follows: 33 * If destination is aligned and source unaligned, 35 * If destination is unaligned, align it by conditionally 70 mov a5, a2 # copy dst so that a2 is return value 73 bbsi.l a2, 0, .Ldst1mod2 # if dst is 1 mod 2 74 bbsi.l a2, 1, .Ldst2mod4 # if dst is 2 mod 4 75 .Ldstaligned: # return here from .Ldstunaligned when dst is aligne [all...] |
/linux-master/arch/x86/crypto/ |
H A D | crc32c-pcl-intel-asm_64.S | 17 * This software is available to you under a choice of one of two 113 ## If len is less than 8 and we're unaligned, we need to jump 219 lea (K_table-8)(%rip), %bufp # first entry is for idx 1 257 # len is less than 8 if we got here 282 jz .Ldo_return # Return if remaining length is zero 292 jz .Ldo_return # return if remaining data is zero 299 jz .Ldo_return # return if remaining data is zero 317 ## jump table Table is 129 entries x 2 bytes each 332 ## Table is 128 entries x 2 words (8 bytes) each
|
/linux-master/tools/tracing/rtla/ |
H A D | Makefile.standalone | 18 $(info This is a deprecated method to compile RTLA, please compile from Linux kernel source)
|
/linux-master/tools/perf/scripts/perl/Perf-Trace-Util/lib/Perf/Trace/ |
H A D | Context.pm | 47 This library is free software; you can redistribute it and/or modify
|