Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add support for microarchitecture levels #31

Merged
merged 5 commits into from
May 20, 2021
Merged

Add support for microarchitecture levels #31

Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
19ea8a6
Add support for microarchitecture levels
pikacic May 11, 2021
cdd91c5
fix gcc version boundaries for psABI x86 microarch levels
pikacic May 12, 2021
d86e243
make microarch feature levels a parallel branch of the DAG
pikacic May 19, 2021
7596ff8
group microarchitecture feature levels
pikacic May 19, 2021
b421338
Fix AMD compatibility chain
pikacic May 20, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
168 changes: 165 additions & 3 deletions cpu/microarchitectures.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -176,8 +176,47 @@
]
}
},
"nehalem": {
"x86_64_v2": {
"from": ["core2"],
"vendor": "generic",
"features": [
"mmx",
"sse",
"sse2",
"ssse3",
"sse4_1",
"sse4_2",
"popcnt"
],
"compilers": {
"gcc": [
{
"versions": "11.1:",
"name": "x86-64-v2",
"flags": "-march={name} -mtune=generic"
},
{
"versions": "4.6:10.3",
Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

which is the correct way of specifying gcc < 11.1? Should I use the latest known 10.x version or a dummy 10.99?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just :11.0 should work. The ranges are (for better or for worse) inclusive.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Done, although GCC 11.0 does not exist (it looks like they have been skipping the x.0 tag recently).
What about Clang?

"name": "x86-64",
"flags": "-march={name} -mtune=generic -mcx16 -msahf -mpopcnt -msse3 -msse4.1 -msse4.2 -mssse3"
}
],
"clang": [
{
"versions": "12.0:",
"name": "x86-64-v2",
"flags": "-march={name} -mtune=generic"
},
{
"versions": "3.9:11.1",
"name": "x86-64",
"flags": "-march={name} -mtune=generic -mcx16 -msahf -mpopcnt -msse3 -msse4.1 -msse4.2 -mssse3"
}
]
}
},
"nehalem": {
"from": ["x86_64_v2"],
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I wonder if:

Suggested change
"from": ["x86_64_v2"],
"from": ["core2", "x86_64_v2"],

is a better modeling choice (and similarly for the other virtual versions below). Need to give some more thought at it.

"vendor": "GenuineIntel",
"features": [
"mmx",
Expand Down Expand Up @@ -393,8 +432,64 @@
]
}
},
"haswell": {
"x86_64_v3": {
"from": ["ivybridge"],
"vendor": "generic",
"features": [
"mmx",
"sse",
"sse2",
"ssse3",
"sse4_1",
"sse4_2",
"popcnt",
"aes",
"pclmulqdq",
"avx",
"rdrand",
"f16c",
"movbe",
"fma",
"avx2",
"bmi1",
"bmi2"
],
"compilers": {
"gcc": [
{
"versions": "11.1:",
"name": "x86-64-v3",
"flags": "-march={name} -mtune=generic"
},
{
"versions": "4.8:10.3",
"name": "x86-64",
"flags": "-march={name} -mtune=generic -mcx16 -msahf -mpopcnt -msse3 -msse4.1 -msse4.2 -mssse3 -mavx -mavx2 -mbmi -mbmi2 -mf16c -mfma -mlzcnt -mmovbe -mxsave"
}
],
"clang": [
{
"versions": "12.0:",
"name": "x86-64-v3",
"flags": "-march={name} -mtune=generic"
},
{
"versions": "3.9:11.1",
"name": "x86-64",
"flags": "-march={name} -mtune=generic -mcx16 -msahf -mpopcnt -msse3 -msse4.1 -msse4.2 -mssse3 -mavx -mavx2 -mbmi -mbmi2 -mf16c -mfma -mlzcnt -mmovbe -mxsave"
}
],
"apple-clang": [
{
"versions": "8.0:",
"name": "x86-64",
"flags": "-march={name} -mtune=generic -mcx16 -msahf -mpopcnt -msse3 -msse4.1 -msse4.2 -mssse3 -mavx -mavx2 -mbmi -mbmi2 -mf16c -mfma -mlzcnt -mmovbe -mxsave"
}
]
}
},
"haswell": {
"from": ["x86_64_v3"],
"vendor": "GenuineIntel",
"features": [
"mmx",
Expand Down Expand Up @@ -641,8 +736,75 @@
]
}
},
"skylake_avx512": {
"x86_64_v4": {
"from": ["skylake"],
"vendor": "generic",
"features": [
"mmx",
"sse",
"sse2",
"ssse3",
"sse4_1",
"sse4_2",
"popcnt",
"aes",
"pclmulqdq",
"avx",
"rdrand",
"f16c",
"movbe",
"fma",
"avx2",
"bmi1",
"bmi2",
"rdseed",
"adx",
"clflushopt",
"xsavec",
"xsaveopt",
"avx512f",
"clwb",
"avx512vl",
"avx512bw",
"avx512dq",
"avx512cd"
],
"compilers": {
"gcc": [
{
"versions": "11.1:",
"name": "x86-64-v4",
"flags": "-march={name} -mtune=generic"
},
{
"versions": "6.0:10.3",
"name": "x86-64",
"flags": "-march={name} -mtune=generic -mcx16 -msahf -mpopcnt -msse3 -msse4.1 -msse4.2 -mssse3 -mavx -mavx2 -mbmi -mbmi2 -mf16c -mfma -mlzcnt -mmovbe -mxsave -mavx512f -mavx512bw -mavx512cd -mavx512dq -mavx512vl"
}
],
"clang": [
{
"versions": "12.0:",
"name": "x86-64-v4",
"flags": "-march={name} -mtune=generic"
},
{
"versions": "3.9:11.1",
"name": "x86-64",
"flags": "-march={name} -mtune=generic -mcx16 -msahf -mpopcnt -msse3 -msse4.1 -msse4.2 -mssse3 -mavx -mavx2 -mbmi -mbmi2 -mf16c -mfma -mlzcnt -mmovbe -mxsave -mavx512f -mavx512bw -mavx512cd -mavx512dq -mavx512vl"
}
],
"apple-clang": [
{
"versions": "8.0:",
"name": "x86-64",
"flags": "-march={name} -mtune=generic -mcx16 -msahf -mpopcnt -msse3 -msse4.1 -msse4.2 -mssse3 -mavx -mavx2 -mbmi -mbmi2 -mf16c -mfma -mlzcnt -mmovbe -mxsave -mavx512f -mavx512bw -mavx512cd -mavx512dq -mavx512vl"
}
]
}
},
"skylake_avx512": {
"from": ["x86_64_v4"],
"vendor": "GenuineIntel",
"features": [
"mmx",
Expand Down