[go-nuts] Re: Call of a Go-function inside C-code
Dear Ian Lance Taylor,
thankyou very much for your hint!
Kind regards, Christian
Am Mittwoch, 2. Oktober 2019 04:36:4D8 UTC+2 schrieb Christian Maurer:
>
> Dear community,
>
> what is wrong with the following simple code, which yields "undefined
> reference for F" ?
>
>
> file f.go:
>
> package main
> // #include "f.h"
> // void f (int a) { F(a) }
> import "C"
>
> func F(a int) { println(a) }
> func main() { C.f(7) }
>
>
> file f.h:
>
> extern void F (int a);
>
> //export F
>
>
> Thanks in advance for any help,
> Christian
>
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/golang-nuts/0b53181a-eb6c-4c0f-a1f9-df5013c0eb55%40googlegroups.com.
[go-nuts] Re: Call of a Go-function inside C-code
Dear Ian Lance Taylor, thankyou very much for your hint! Kind regards, Christian -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/969b21e9-7b90-4e39-bb63-2f9a6418d198%40googlegroups.com.
[go-nuts] Call of a Go-function inside C-code
Dear community,
what is wrong with the following simple code, which yields "undefined
reference for F" ?
file f.go:
package main
// #include "f.h"
// void f (int a) { F(a) }
import "C"
func F(a int) { println(a) }
func main() { C.f(7) }
file f.h:
extern void F (int a);
//export F
Thanks in advance for any help,
Christian
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/golang-nuts/e31defe4-fde3-47c5-b052-8979939257a2%40googlegroups.com.
[go-nuts] Re: mutual exclusion algorithm of Dijkstra - strange behaviour
Dear Community and dear Go-developers,
Meanwhile it is clear why things do not work:
The Go-Scheduler is unable to allow to switch to another goroutine in
busy-waiting-loops -
the only possibility to get around that problem is either to put
"switch-steps" into the source
- either "time.Sleep(1)" or "runtime.Gosched()".
I think that THIS SHOULD BE DOCUMENTED IN THE LANGUAGE SPECIFICATION !
(See my book on https://maurer-berlin.eu/nspbuch - in it's 4th edition
things are mentioned.)
With kind regards,
Christian Maurer
Am Montag, 30. Oktober 2017 18:30:06 UTC+1 schrieb Christian Maurer:
>
> Dear Go-community,
>
> I noticed a very strange effect by translating the
> mutual exclusion algorithm of E. W. Dijkstra to Go.
> Reference:
> Cooperating Sequential Processes. Technical Report EWD-123,
> Technological University Eindhoven (1965)
> http://www.cs.utexas.edu/users/EWD/ewd01xx/EWD123.PDF)
>
> Inserting or omitting code lines A and B resp.
> changes the behaviour of this program very severely:
>
> With line A and without line B:
> mutual exclusion is not guaranteed,
> with line B and without line A:
> lock does not terminate and
> with both lines:
> the program works as it should.
>
> --- 8<
> package main
>
> import (
> "math/rand"
> "time"
> )
>
> const N = 10 // number of goroutines involved
>
> var (
> turn int
> b, c = make([]int, N+1), make([]int, N+1)
> n= uint(0)
> inCS = make([]bool, N+1)
> s= make([]string, N+1)
> done = make(chan int)
> )
>
> func chk(i int) {
> for j := 1; j <= N; j++ {
> if j != i && inCS[j] {
> panic("oops")
> }
> }
> }
>
> func lock(i int) {
> b[i] = 0
> L:
> if turn != i {
> c[i] = 1
> if b[turn] == 1 {
> turn = i
> goto L
> }
> }
> time.Sleep(1) // A
> c[i] = 0
> time.Sleep(1) // B
> for j := 1; j <= N; j++ {
> if j != i && c[j] == 0 {
> goto L
> }
> }
> inCS[i] = true
> chk(i)
> }
>
> func unlock(i int) {
> inCS[i] = false
> c[i], b[i] = 1, 1
> // turn = 0 // does not matter
> }
>
> func v() {
> time.Sleep(time.Duration(1e7 + rand.Int63n(1e7)))
> }
>
> func count(i int) {
> for k := 0; k < 100; k++ {
> lock(i)
> accu := n
> v()
> accu++
> v()
> n = accu
> v()
> println(s[i], n)
> unlock(i)
> v()
> }
> done <- 0
> }
>
> func main() {
> s[1] = ""
> for i := 2; i <= N; i++ {
> s[i] = s[i-1] + " "
> }
> for i := 1; i <= N; i++ {
> go count(i)
> }
> for i := 1; i <= N; i++ {
> <-done
> }
> }
> --- >8
>
> The cause for the "freaking out" of the program without
> those compulsory breaks (of only 1 ns) is absolutely unclear.
>
> Any sort of help is welcome!
>
> Kind regards to everybody,
>
> Christian Maurer
>
>
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/golang-nuts/f83856a2-0661-4425-9861-196b83fcf07f%40googlegroups.com.
[go-nuts] mutual exclusion algorithm of Dijkstra - strange behaviour
Dear Go-community,
I noticed a very strange effect by translating the
mutual exclusion algorithm of E. W. Dijkstra to Go.
Reference:
Cooperating Sequential Processes. Technical Report EWD-123,
Technological University Eindhoven (1965)
http://www.cs.utexas.edu/users/EWD/ewd01xx/EWD123.PDF)
Inserting or omitting code lines A and B resp.
changes the behaviour of this program very severely:
With line A and without line B:
mutual exclusion is not guaranteed,
with line B and without line A:
lock does not terminate and
with both lines:
the program works as it should.
--- 8<
package main
import (
"math/rand"
"time"
)
const N = 10 // number of goroutines involved
var (
turn int
b, c = make([]int, N+1), make([]int, N+1)
n= uint(0)
inCS = make([]bool, N+1)
s= make([]string, N+1)
done = make(chan int)
)
func chk(i int) {
for j := 1; j <= N; j++ {
if j != i && inCS[j] {
panic("oops")
}
}
}
func lock(i int) {
b[i] = 0
L:
if turn != i {
c[i] = 1
if b[turn] == 1 {
turn = i
goto L
}
}
time.Sleep(1) // A
c[i] = 0
time.Sleep(1) // B
for j := 1; j <= N; j++ {
if j != i && c[j] == 0 {
goto L
}
}
inCS[i] = true
chk(i)
}
func unlock(i int) {
inCS[i] = false
c[i], b[i] = 1, 1
// turn = 0 // does not matter
}
func v() {
time.Sleep(time.Duration(1e7 + rand.Int63n(1e7)))
}
func count(i int) {
for k := 0; k < 100; k++ {
lock(i)
accu := n
v()
accu++
v()
n = accu
v()
println(s[i], n)
unlock(i)
v()
}
done <- 0
}
func main() {
s[1] = ""
for i := 2; i <= N; i++ {
s[i] = s[i-1] + " "
}
for i := 1; i <= N; i++ {
go count(i)
}
for i := 1; i <= N; i++ {
<-done
}
}
--- >8
The cause for the "freaking out" of the program without
those compulsory breaks (of only 1 ns) is absolutely unclear.
Any sort of help is welcome!
Kind regards to everybody,
Christian Maurer
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
dijkstra.go
Description: Binary data
