Debian Codegraph @

upstream/debian-unstable/man3/OPENSSL_ia32cap.3ssl

raw
1 .\" -*- mode: troff; coding: utf-8 -*-
2 .\" Automatically generated by Pod::Man 5.0102 (Pod::Simple 3.45)
3 .\"
4 .\" Standard preamble:
5 .\" ========================================================================
6 .de Sp \" Vertical space (when we can't use .PP)
7 .if t .sp .5v
8 .if n .sp
9 ..
10 .de Vb \" Begin verbatim text
11 .ft CW
12 .nf
13 .ne \\$1
14 ..
15 .de Ve \" End verbatim text
16 .ft R
17 .fi
18 ..
19 .\" \*(C` and \*(C' are quotes in nroff, nothing in troff, for use with C<>.
20 .ie n \{\
21 . ds C` ""
22 . ds C' ""
23 'br\}
24 .el\{\
25 . ds C`
26 . ds C'
27 'br\}
28 .\"
29 .\" Escape single quotes in literal strings from groff's Unicode transform.
30 .ie \n(.g .ds Aq \(aq
31 .el .ds Aq '
32 .\"
33 .\" If the F register is >0, we'll generate index entries on stderr for
34 .\" titles (.TH), headers (.SH), subsections (.SS), items (.Ip), and index
35 .\" entries marked with X<> in POD. Of course, you'll have to process the
36 .\" output yourself in some meaningful fashion.
37 .\"
38 .\" Avoid warning from groff about undefined register 'F'.
39 .de IX
40 ..
41 .nr rF 0
42 .if \n(.g .if rF .nr rF 1
43 .if (\n(rF:(\n(.g==0)) \{\
44 . if \nF \{\
45 . de IX
46 . tm Index:\\$1\t\\n%\t"\\$2"
47 ..
48 . if !\nF==2 \{\
49 . nr % 0
50 . nr F 2
51 . \}
52 . \}
53 .\}
54 .rr rF
55 .\" ========================================================================
56 .\"
57 .IX Title "OPENSSL_IA32CAP 3SSL"
58 .TH OPENSSL_IA32CAP 3SSL 2026-04-07 3.6.2 OpenSSL
59 .\" For nroff, turn off justification. Always turn off hyphenation; it makes
60 .\" way too many mistakes in technical documents.
61 .if n .ad l
62 .nh
63 .SH NAME
64 OPENSSL_ia32cap \- the x86[_64] processor capabilities vector
65 .SH SYNOPSIS
66 .IX Header "SYNOPSIS"
67 .Vb 1
68 \& env OPENSSL_ia32cap=... <application>
69 .Ve
70 .SH DESCRIPTION
71 .IX Header "DESCRIPTION"
72 OpenSSL supports a range of x86[_64] instruction set extensions and
73 features. These extensions are denoted by individual bits or groups of bits
74 stored internally as ten 32\-bit capability vectors and for simplicity
75 represented logically below as five 64\-bit vectors. This logical
76 vector (LV) representation is used to streamline the definition of the
77 OPENSSL_ia32cap environment variable.
78 .PP
79 Upon toolkit initialization, the capability vectors are populated through
80 successive executions of the CPUID instruction, after which any OPENSSL_ia32cap
81 environment variable capability bit modifications are applied. After toolkit
82 initialization is complete, populated vectors are then used to choose
83 between different code paths to provide optimal performance across a wide
84 range of x86[_64] based processors.
85 .PP
86 Further CPUID information can be found in the Intel(R) Architecture
87 Instruction Set Extensions Programming Reference, and the AMD64 Architecture
88 Programmer's Manual (Volume 3).
89 .SS "Notable Capability Bits for LV0"
90 .IX Subsection "Notable Capability Bits for LV0"
91 The following are notable capability bits from logical vector 0 (LV0)
92 resulting from the following execution of CPUID.(EAX=01H).EDX and
93 CPUID.(EAX=01H).ECX:
94 .IP "bit #0+4 denoting presence of Time-Stamp Counter;" 4
95 .IX Item "bit #0+4 denoting presence of Time-Stamp Counter;"
96 .PD 0
97 .IP "bit #0+19 denoting availability of CLFLUSH instruction;" 4
98 .IX Item "bit #0+19 denoting availability of CLFLUSH instruction;"
99 .IP "bit #0+20, reserved by Intel, is used to choose among RC4 code paths;" 4
100 .IX Item "bit #0+20, reserved by Intel, is used to choose among RC4 code paths;"
101 .IP "bit #0+23 denoting MMX support;" 4
102 .IX Item "bit #0+23 denoting MMX support;"
103 .IP "bit #0+24, FXSR bit, denoting availability of XMM registers;" 4
104 .IX Item "bit #0+24, FXSR bit, denoting availability of XMM registers;"
105 .IP "bit #0+25 denoting SSE support;" 4
106 .IX Item "bit #0+25 denoting SSE support;"
107 .IP "bit #0+26 denoting SSE2 support;" 4
108 .IX Item "bit #0+26 denoting SSE2 support;"
109 .IP "bit #0+28 denoting Hyperthreading, which is used to distinguish cores with shared cache;" 4
110 .IX Item "bit #0+28 denoting Hyperthreading, which is used to distinguish cores with shared cache;"
111 .IP "bit #0+30, reserved by Intel, denotes specifically Intel CPUs;" 4
112 .IX Item "bit #0+30, reserved by Intel, denotes specifically Intel CPUs;"
113 .IP "bit #0+33 denoting availability of PCLMULQDQ instruction;" 4
114 .IX Item "bit #0+33 denoting availability of PCLMULQDQ instruction;"
115 .IP "bit #0+41 denoting SSSE3, Supplemental SSE3, support;" 4
116 .IX Item "bit #0+41 denoting SSSE3, Supplemental SSE3, support;"
117 .IP "bit #0+43 denoting AMD XOP support (forced to zero on non-AMD CPUs);" 4
118 .IX Item "bit #0+43 denoting AMD XOP support (forced to zero on non-AMD CPUs);"
119 .IP "bit #0+54 denoting availability of MOVBE instruction;" 4
120 .IX Item "bit #0+54 denoting availability of MOVBE instruction;"
121 .IP "bit #0+57 denoting AES-NI instruction set extension;" 4
122 .IX Item "bit #0+57 denoting AES-NI instruction set extension;"
123 .IP "bit #0+58, XSAVE bit, lack of which in combination with MOVBE is used to identify Atom Silvermont core;" 4
124 .IX Item "bit #0+58, XSAVE bit, lack of which in combination with MOVBE is used to identify Atom Silvermont core;"
125 .IP "bit #0+59, OSXSAVE bit, denoting availability of YMM registers;" 4
126 .IX Item "bit #0+59, OSXSAVE bit, denoting availability of YMM registers;"
127 .IP "bit #0+60 denoting AVX extension;" 4
128 .IX Item "bit #0+60 denoting AVX extension;"
129 .IP "bit #0+62 denoting availability of RDRAND instruction;" 4
130 .IX Item "bit #0+62 denoting availability of RDRAND instruction;"
131 .PD
132 .SS "Notable Capability Bits for LV1"
133 .IX Subsection "Notable Capability Bits for LV1"
134 The following are notable capability bits from logical vector 1 (LV1)
135 resulting from the following execution of CPUID.(EAX=07H,ECX=0H).EBX and
136 CPUID.(EAX=07H,ECX=0H).ECX:
137 .IP "bit #64+3 denoting availability of BMI1 instructions, e.g. ANDN;" 4
138 .IX Item "bit #64+3 denoting availability of BMI1 instructions, e.g. ANDN;"
139 .PD 0
140 .IP "bit #64+5 denoting availability of AVX2 instructions;" 4
141 .IX Item "bit #64+5 denoting availability of AVX2 instructions;"
142 .IP "bit #64+8 denoting availability of BMI2 instructions, e.g. MULX and RORX;" 4
143 .IX Item "bit #64+8 denoting availability of BMI2 instructions, e.g. MULX and RORX;"
144 .IP "bit #64+16 denoting availability of AVX512F extension;" 4
145 .IX Item "bit #64+16 denoting availability of AVX512F extension;"
146 .IP "bit #64+17 denoting availability of AVX512DQ extension;" 4
147 .IX Item "bit #64+17 denoting availability of AVX512DQ extension;"
148 .IP "bit #64+18 denoting availability of RDSEED instruction;" 4
149 .IX Item "bit #64+18 denoting availability of RDSEED instruction;"
150 .IP "bit #64+19 denoting availability of ADCX and ADOX instructions;" 4
151 .IX Item "bit #64+19 denoting availability of ADCX and ADOX instructions;"
152 .IP "bit #64+21 denoting availability of AVX512IFMA extension;" 4
153 .IX Item "bit #64+21 denoting availability of AVX512IFMA extension;"
154 .IP "bit #64+29 denoting availability of SHA extension;" 4
155 .IX Item "bit #64+29 denoting availability of SHA extension;"
156 .IP "bit #64+30 denoting availability of AVX512BW extension;" 4
157 .IX Item "bit #64+30 denoting availability of AVX512BW extension;"
158 .IP "bit #64+31 denoting availability of AVX512VL extension;" 4
159 .IX Item "bit #64+31 denoting availability of AVX512VL extension;"
160 .IP "bit #64+41 denoting availability of VAES extension;" 4
161 .IX Item "bit #64+41 denoting availability of VAES extension;"
162 .IP "bit #64+42 denoting availability of VPCLMULQDQ extension;" 4
163 .IX Item "bit #64+42 denoting availability of VPCLMULQDQ extension;"
164 .PD
165 .SS "Notable Capability Bits for LV2"
166 .IX Subsection "Notable Capability Bits for LV2"
167 The following are notable capability bits from logical vector 2 (LV2)
168 resulting from the following execution of CPUID.(EAX=07H,ECX=0H).EDX and
169 CPUID.(EAX=07H,ECX=1H).EAX:
170 .IP "bit #128+15 denoting availability of Hybrid CPU;" 4
171 .IX Item "bit #128+15 denoting availability of Hybrid CPU;"
172 .PD 0
173 .IP "bit #128+29 denoting support for IA32_ARCH_CAPABILITIES MSR;" 4
174 .IX Item "bit #128+29 denoting support for IA32_ARCH_CAPABILITIES MSR;"
175 .IP "bit #128+32 denoting availability of SHA512 extension;" 4
176 .IX Item "bit #128+32 denoting availability of SHA512 extension;"
177 .IP "bit #128+33 denoting availability of SM3 extension;" 4
178 .IX Item "bit #128+33 denoting availability of SM3 extension;"
179 .IP "bit #128+34 denoting availability of SM4 extension;" 4
180 .IX Item "bit #128+34 denoting availability of SM4 extension;"
181 .IP "bit #128+55 denoting availability of AVX-IFMA extension;" 4
182 .IX Item "bit #128+55 denoting availability of AVX-IFMA extension;"
183 .PD
184 .SS "Notable Capability Bits for LV3"
185 .IX Subsection "Notable Capability Bits for LV3"
186 The following are notable capability bits from logical vector 3 (LV3)
187 resulting from the following execution of CPUID.(EAX=07H,ECX=1H).EDX and
188 CPUID.(EAX=07H,ECX=1H).EBX:
189 .IP "bit #192+19 denoting availability of AVX10 Converged Vector ISA extension;" 4
190 .IX Item "bit #192+19 denoting availability of AVX10 Converged Vector ISA extension;"
191 .PD 0
192 .IP "bit #192+21 denoting availability of APX_F extension;" 4
193 .IX Item "bit #192+21 denoting availability of APX_F extension;"
194 .PD
195 .SS "Notable Capability Bits for LV4"
196 .IX Subsection "Notable Capability Bits for LV4"
197 The following are notable capability bits from logical vector 4 (LV4)
198 resulting from the following execution of CPUID.(EAX=07H,ECX=1H).ECX and
199 CPUID.(EAX=24H,ECX=0H).EBX:
200 .IP "bits #256+32+[0:7] denoting AVX10 Converged Vector ISA Version (8 bits);" 4
201 .IX Item "bits #256+32+[0:7] denoting AVX10 Converged Vector ISA Version (8 bits);"
202 .PD 0
203 .IP "bit #256+48 denoting AVX10 XMM support;" 4
204 .IX Item "bit #256+48 denoting AVX10 XMM support;"
205 .IP "bit #256+49 denoting AVX10 YMM support;" 4
206 .IX Item "bit #256+49 denoting AVX10 YMM support;"
207 .IP "bit #256+50 denoting AVX10 ZMM support;" 4
208 .IX Item "bit #256+50 denoting AVX10 ZMM support;"
209 .PD
210 .SS "OPENSSL_ia32cap environment variable"
211 .IX Subsection "OPENSSL_ia32cap environment variable"
212 The \fBOPENSSL_ia32cap\fR environment variable provides a mechanism to override
213 the default capability vector values at library initialization time.
214 The variable consists of a series of 64\-bit numbers representing each
215 of the logical vectors (LV) described above. Each value is delimited by a '\fB:\fR'.
216 Decimal/Octal/Hexadecimal values representations are supported.
217 .PP
218 \&\f(CW\*(C`env OPENSSL_ia32cap=LV0:LV1:LV2:LV3:LV4\*(C'\fR
219 .PP
220 Used in this form, each non-null logical vector will *overwrite* the entire corresponding
221 capability vector pair with the provided value. To keep compatibility with the
222 behaviour of the original OPENSSL_ia32cap environment variable
223 <env OPENSSL_ia32cap=LV0:LV1>, the next capability vector pairs will be set to zero.
224 .PP
225 To illustrate, the following will zero all capability bits in logical vectors 1 and further
226 (disable all post-AVX extensions):
227 .PP
228 \&\f(CW\*(C`env OPENSSL_ia32cap=:0\*(C'\fR
229 .PP
230 The following will zero all capability bits in logical vectors 2 and further:
231 .PP
232 \&\f(CW\*(C`env OPENSSL_ia32cap=::0\*(C'\fR
233 .PP
234 The following will zero all capability bits only in logical vector 1:
235 \&\f(CW\*(C`env OPENSSL_ia32cap=:0::::\*(C'\fR
236 .PP
237 A more likely usage scenario would be to disable specific instruction set extensions.
238 The '\fB~\fR' character is used to specify a bit mask of the extensions to be disabled for
239 a particular logical vector.
240 .PP
241 To illustrate, the following will disable AVX2 code paths and further extensions:
242 .PP
243 \&\f(CW\*(C`env OPENSSL_ia32cap=:~0x20000000000\*(C'\fR
244 .PP
245 The following will disable AESNI (LV0 bit 57) and VAES (LV1 bit 41)
246 extensions and therefore any code paths using those extensions but leave
247 the rest of the logical vectors unchanged:
248 .PP
249 \&\f(CW\*(C`env OPENSSL_ia32cap=~0x200000000000000:~0x20000000000:~0x0:~0x0:~0x0\*(C'\fR
250 .SH NOTES
251 .IX Header "NOTES"
252 Not all capability bits are copied from CPUID output verbatim. An example
253 of this is the somewhat less intuitive clearing of LV0 bit #28, or ~0x10000000
254 in the "environment variable" terms. It has been adjusted to reflect whether or
255 not the data cache is actually shared between logical cores. This in turn affects
256 the decision on whether or not expensive countermeasures against cache-timing attacks
257 are applied, most notably in AES assembler module.
258 .SH "RETURN VALUES"
259 .IX Header "RETURN VALUES"
260 Not available.
261 .SH COPYRIGHT
262 .IX Header "COPYRIGHT"
263 Copyright 2004\-2021 The OpenSSL Project Authors. All Rights Reserved.
264 .PP
265 Licensed under the Apache License 2.0 (the "License"). You may not use
266 this file except in compliance with the License. You can obtain a copy
267 in the file LICENSE in the source distribution or at
268 <https://www.openssl.org/source/license.html>.