This PR adds support for reading the frequency of Apple Silicon
M1/M2 CPUs. We do so by reading the values out of the IOKit
framework, as a few other projects have now demonstrated to be
possible. This requires the use of CGO. The library provides a
convenience IsAppleSilicon() guard to detect whether the values
can be read.
Currently gopsutil does not support the big.LITTLE CPU architectures
(i think?) - in fact the P and E cores have different max frequencies.
For now, just read the P core frequency. The E core data is readily
available if we want to read it in the future.
Closes#1000
Small example program
```go
package main
import (
"fmt"
"github.com/shoenig/go-m1cpu"
"github.com/shirou/gopsutil/v3/cpu"
)
func main() {
fmt.Println("is Apple Silicon:", m1cpu.IsAppleSilicon())
fmt.Println("model name", m1cpu.ModelName())
fmt.Println("pCore GHz", m1cpu.PCoreGHz())
fmt.Println("eCore GHz", m1cpu.ECoreGHz())
fmt.Println("pCore Hz", m1cpu.PCoreHz())
fmt.Println("eCore Hz", m1cpu.ECoreHz())
fmt.Println("----- gopsutil ----")
infos, err := cpu.Info()
if err != nil {
panic(err)
}
for _, info := range infos {
fmt.Println("info.Mhz", info.Mhz)
}
}
```
```shell
go run main.go
is Apple Silicon: true
model name Apple M2 Pro
pCore GHz 3.504
eCore GHz 2.424
pCore Hz 3504000000
eCore Hz 2424000000
----- gopsutil ----
info.Mhz 3.504e+09
```
This commit replaces `os.Setenv` with `t.Setenv` in tests. The
environment variable is automatically restored to its original value
when the test and all its subtests complete.
Reference: https://pkg.go.dev/testing#T.Setenv
Signed-off-by: Eng Zer Jun <engzerjun@gmail.com>
Use GetActiveProcessorCount and the ALL_PROCESSOR_GROUPS const provided
in golang.org/x/sys/windows. The function is available on Windows 7 and
later. Go requires Windows 7, see https://go.dev/doc/go1.11#ports
Even thought OpenBSD often breaks the ABI compatibility and doesn't make
*any* promise of "stability", this project aims to be "pure go" so avoid
doing inter-op at the cost of artificially reducing the number of
supported architectures down to amd64 and i386.
To add support for another architecture (e.g. arm), add another file
cpu_openbsd_${arch}.go like done for 386 and amd64. The fields are
declared as `long' in C, so pick the appropriate size when declaring the
struct.
don't make assumptions on which CPUs are online and wich aren't based
on hw.smt and hw.ncpuonline. Rather, use KERN_CPUSTATS to get the CPU
statistics, which includes a flag field that can tell us if that CPU
is online or not.
We can't use unix.Sysctl* for some sysctls, so we're on our own with
converting data from C arrays.
Don't assume that the byte order is little endian but do the right
thing. Moreover, there's a little distinction in the sizes reported
by KERN_CPTIME (long[cpustates]) and KERN_CPTIME2
(u_int64_t[cpustates]) so account for that too.
updated win32_Processor struct to exclude loadpercentage field.
The loadpercentage takes linearly more time as the # of sockets
increases. By default vSphere maps 1 vCPU to 1 socket, resulting in very
poor performance when getting CPU info against, saying, 40 vCPU VM
(basically 40 sockets as seen by the VM).
Currently, ClocksPerSec is determined by exec'ing getconf in func init,
i.e. on startup of every program importing the package. getconf might
not be present on some systems or is not executable by the current user.
To avoid this hard to control dependency, use the
github.com/tklauser/go-sysconf package which implements sysconf(3)
entirely in Go without cgo. The package is supported on all platforms
currently supported by the cpu and v3/cpu package of gopsutil.
