Re: Re[2]: [9fans] Strange 'uintptr' behavior
Ah, thanks! Going to try... út 25. 1. 2022 v 9:28 odesílatel Alex Musolino napsal: > It didn't always work on 9front. Check the dump! > > After bisecting the dump and looking through the git log, I suspect > that 9front changeset 71939a82 [1] will fix it for you. > > [1] > https://git.9front.org/plan9front/plan9front/71939a82ccc505b0baa990d7fe17e7085d0ab8fc/commit.html > -- 9fans: 9fans Permalink: https://9fans.topicbox.com/groups/9fans/Tb5fe0dc43447ab2c-Ma557fc25c99094fa66c5fa53 Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
Re: [9fans] Strange 'uintptr' behavior
Sorry, I was not precise enough.
I prepared arrays in the initiated data segment:
This program cannot be compiled:
#include
#include
uintptr frst[] = {nil, nil, nil, nil, nil};
uintptr scnd[] = {nil, nil, frst, &frst[1]};
void main(void) {
exits(nil);
}
...because of:
cpu% 5c t.c
t.c:5 initializer is not a constant: scnd
This program can be compiled:
#include
#include
void* frst[] = {nil, nil, nil, nil, nil};
void* scnd[] = {nil, nil, frst, &frst[1]};
void main(void) {
exits(nil);
}
Pavel
--
9fans: 9fans
Permalink:
https://9fans.topicbox.com/groups/9fans/Tb5fe0dc43447ab2c-M55c6dc11d2315bac025e3124
Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
Re: [9fans] Strange 'uintptr' behavior
> what headers you've included? > As usually: #include #include with uintptr i get compilation error: incompatible types: "ULONG" and "IND > VOID" for op AS. > with void* it compiles without errors. > > alex.b. > > ps: are you czech? > Yes. Pavel -- 9fans: 9fans Permalink: https://9fans.topicbox.com/groups/9fans/Tb5fe0dc43447ab2c-Mfd55c3e288c7cbeeafb039f5 Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
[9fans] Strange 'uintptr' behavior
Hi all,
I tried some artificial example for Plan9 compiler (5c)...
This code is compiled without any problem:
uintptr frst[] = {nil, nil, nil, nil, nil};
uintptr scnd[] = {nil, nil, frst, &frst[0]};
But this one fails:
uintptr frst[] = {nil, nil, nil, nil, nil};
uintptr scnd[] = {nil, nil, frst, &frst[1]};
...with message:
cpu% 5c -p t.c
/usr/pavel/t.c:16[t.c:888] initializer is not a constant: scnd
OTOH, replacement of 'uintptr' by 'void*' is compiled without any problem:
void* frst[] = {nil, nil, nil, nil, nil};
void* scnd[] = {nil, nil, frst, &frst[1]};
Any explanation is appreciated.
Pavel
--
9fans: 9fans
Permalink:
https://9fans.topicbox.com/groups/9fans/Tb5fe0dc43447ab2c-M240dd4253f2ea03efc2d9013
Delivery options: https://9fans.topicbox.com/groups/9fans/subscription
Re: [9fans] the fossil (tm) stolen
> I Really use fossil and venti and have done so for the last 12 years, > though perhaps I am nobody... > So we are 2 nobodies, at least... ;) Pavel
[9fans] Strange behavior of 'go's' DialUDP under Plan9
Hi all,
I am trying a very simple 'go' program under Plan9 (both 386 and ARM):
package main
import (
"fmt"
"net"
"time"
)
func main() {
laddr, err := net.ResolveUDPAddr("udp", ":6000")
if err != nil {
fmt.Println("laddr: ", err)
}
raddr, err := net.ResolveUDPAddr("udp", "cpx61.tk8.lan:6000")
if err != nil {
fmt.Println("raddr: ", err)
}
conn, err := net.DialUDP("udp", laddr, raddr)
if err != nil {
fmt.Println("Error: ", err)
}
defer conn.Close()
time.Sleep(time.Minute)
}
When running I check 'netstat' listing:
udp 0pavel Open 39505 6000 cpx61.tk8.lan
It seems 'go' port for Plan9 incorrectly handles local port specification
in DialUDP function.
Pavel
Re: [9fans] Change of plan9 partition size
> > I just created a new partition of PLAN9 type there (according to >> https://9p.io/wiki/plan9/setting_up_Venti/index.html), saved and >> rebooted... >> >> ...and system failed to boot: >> >> boot: can't connect to file server: '/boot/kfs' does not exists >> > > It seems the boot process does not like two PLAN9 partitions on SD. > I was wrong. The problem is in disk/fdisk which probably damaged the partition table disabling subsequent reboot. If I partitioned SD under Linux, the problem was over... Pavel
Re: [9fans] Change of plan9 partition size
> > I just created a new partition of PLAN9 type there (according to > https://9p.io/wiki/plan9/setting_up_Venti/index.html), saved and > rebooted... > > ...and system failed to boot: > > boot: can't connect to file server: '/boot/kfs' does not exists > It seems the boot process does not like two PLAN9 partitions on SD. Not clear why... Pavel
Re: [9fans] Change of plan9 partition size
> > > Maybe it would be wise to use Venti on that unsed space... ;) > > If you don't have a venti server somewhere near your pi yet, then > definitely yes. > It will be a fight... ;) I just created a new partition of PLAN9 type there (according to https://9p.io/wiki/plan9/setting_up_Venti/index.html), saved and rebooted... ...and system failed to boot: boot: can't connect to file server: '/boot/kfs' does not exists Pavel
Re: [9fans] Change of plan9 partition size
> > 9pi has about 1,7GB plan9 partition size. > On my 8GB SD card the rest is unused... > Maybe it would be wise to use Venti on that unsed space... ;) Pavel
[9fans] Change of plan9 partition size
> > Once you have that, you can run it on Plan 9 on the Raspberry Pi to > bootstrap the current release of go. > Another stupid question - sorry for that ;)... 9pi has about 1,7GB plan9 partition size. On my 8GB SD card the rest is unused... Is there a way how to extended plan9 partition to use the whole SD space? Thanks in advance. Pavel
Re: [9fans] Go 1.4.3 compilation on Raspberry Pi
> > There are some binaries available here if you want to use them to > bootstrap: > > http://www.9legacy.org/download.html Already done and working fine. ;) Thank you. Pavel
Re: [9fans] Go 1.4.3 compilation on Raspberry Pi
> > I believe that route to bootstrapping go from scratch on Plan 9 > will work only for 386. > I see. > On arm, you can either cross-compile go1.4 on another go platform > (eg plan9/386, plan9port on linux/386, or linux/arm), or start with > a pre-compiled plan9/arm package (there are several to choose from > on http://www.9legacy.org/download.html). > Excellent. Once you have that, you can run it on Plan 9 on the Raspberry Pi to > bootstrap the current release of go. > Perfect, going to follow your recommendation. Thank you for such a hint Pavel
[9fans] Go 1.4.3 compilation on Raspberry Pi
Hi all, I am trying to compile Go 1.4.3 on my Raspberry Pi following David's instructions on https://github.com/golang/go/wiki/Plan9. But still having errors... term% make.rc # Building C bootstrap tool. cmd/dist # Building compilers and Go bootstrap tool for host, plan9/arm. lib9 libbio liblink /usr/glenda/go1.4/include/plan9/../link.h:56[/usr/glenda/go1.4/src/liblink/anames6.c:1487] syntax error, last name: float64 /usr/glenda/go1.4/include/plan9/../link.h:56[/usr/glenda/go1.4/src/liblink/anames8.c:1487] syntax error, last name: float64 go tool dist: FAILED: /bin/5c -FTVwp -DPLAN9 -D__STDC__=1 -D__SIZE_TYPE__=ulong -I/usr/glenda/go1.4/include/plan9 -I/usr/glenda/go1.4/include/plan9/arm -I /usr/glenda/go1.4/src/liblink -o $WORK/anames6.5 /usr/glenda/go1.4/src/liblink/anames6.c: '/usr/glenda/go1.4/src/liblink/anames8.c' does not exist /usr/glenda/go1.4/include/plan9/../link.h:56[/usr/glenda/go1.4/src/liblink/asm5.c:1519] syntax error, last name: float64 go tool dist: FAILED: /bin/5c -FTVwp -DPLAN9 -D__STDC__=1 -D__SIZE_TYPE__=ulong -I/usr/glenda/go1.4/include/plan9 -I/usr/glenda/go1.4/include/plan9/arm -I /usr/glenda/go1.4/src/liblink -o $WORK/anames8.5 /usr/glenda/go1.4/src/liblink/anames8.c: '/usr/glenda/go1.4/src/liblink/anames8.c' does not exist /usr/glenda/go1.4/include/plan9/../link.h:56[/usr/glenda/go1.4/src/liblink/anames5.c:1487] syntax error, last name: float64 go tool dist: FAILED: /bin/5c -FTVwp -DPLAN9 -D__STDC__=1 -D__SIZE_TYPE__=ulong -I/usr/glenda/go1.4/include/plan9 -I/usr/glenda/go1.4/include/plan9/arm -I /usr/glenda/go1.4/src/liblink -o $WORK/asm5.5 /usr/glenda/go1.4/src/liblink/asm5.c: '/usr/glenda/go1.4/src/liblink/anames8.c' does not exist go tool dist: FAILED: /bin/5c -FTVwp -DPLAN9 -D__STDC__=1 -D__SIZE_TYPE__=ulong -I/usr/glenda/go1.4/include/plan9 -I/usr/glenda/go1.4/include/plan9/arm -I /usr/glenda/go1.4/src/liblink -o $WORK/anames5.5 /usr/glenda/go1.4/src/liblink/anames5.c: '/usr/glenda/go1.4/src/liblink/anames8.c' does not exist term% Any idea what I am doing wrong? Thanks in advance. Pavel
Re: [9fans] Plan9 CPU on Rpi has nvram problem
As expected, it was my fault. I did not... cp config-cpu.txt config.txt Sorry for the noise... Pavel
Re: [9fans] Plan9 CPU on Rpi has nvram problem
> > Have you written your key? (hint: auth/wrkey) Still the same: cpu% auth/wrkey readnvram: can't find nvram authid: bootes ... can't write key to nvram: error writing nvram: cpu% The nvram partition exists: /dev/sdM0/nvram Pavel
[9fans] Plan9 CPU on Rpi has nvram problem
Hello everyone, I almost succussfully installed P9 CPU server on my Raspberry Pi. However during the booting it has a problem with nvram each time: 496M memory: 101M kernel data, 395M user, 1877M swap usb/hub... usb/ether... etherusb smsc: b827eb19b423 usb/kb... usb/kb... readnvram: can't find nvram authid: bootes ... Maybe I made a mistake during the CPU server configuration, but cannot find which one... :( Can you help me? Thanks in advance. Pavel
Re: [9fans] 9atom boot problem
> > If I choose to install plan9, I get a result shown in the attached > a.png (and it does not proceed any further). > > If I choose to just boot (not install), it gets stuck after > init: starting /bin/rc > and shows the same 'spurious interrupt 39'. > Try to play with AHCI settings in BIOS. IIRC I had similar problems with my IntelAtom based M/B. Pavel
Re: [9fans] Doesn't 'close' call finish pending 'read' on the same 'fd'?
> + if(strcmp(f[0], "hangup") == 0){
>> + qhangup(c->rq, nil);
>> + qhangup(c->wq, nil);
>> + return nil;
>> + }
>
>
You patch works as expected. ;)
Thanks.
Pavel
Re: [9fans] Doesn't 'close' call finish pending 'read' on the same 'fd'?
> the changes to implement udp hangup are trivial tho:
>
> diff -r 51564dc1adae sys/src/9/ip/udp.c
> --- a/sys/src/9/ip/udp.cThu Nov 13 10:23:53 2014 +0100
> +++ b/sys/src/9/ip/udp.cThu Nov 13 15:42:24 2014 +0100
> @@ -518,6 +518,11 @@
>
> ucb = (Udpcb*)c->ptcl;
> if(n == 1){
> + if(strcmp(f[0], "hangup") == 0){
> + qhangup(c->rq, nil);
> + qhangup(c->wq, nil);
> + return nil;
> + }
>
I see.
Going to try.
Pavel
Re: [9fans] Doesn't 'close' call finish pending 'read' on the same 'fd'?
> interesting, that might be worth implementing. > Maybe. > the alternative would be to send the reading process a note to > interrupt the blocking syscall. > Actually, it my 'plan B'... ;) the closer proc should just send the note and not close the fd. > and the reader proc should close the fd once he's out of the > reading loop. > Exactly! Well, I would prefer a solution with 'hangup', it seems cleaner to me... ...but since it is not implemented in UDP driver, I will implement the note based solution. Pavel Pavel
Re: [9fans] Doesn't 'close' call finish pending 'read' on the same 'fd'?
OTOH, 'hangup' approach does not work in UDP. It means, read() is not finished... :( Pavel 2014-11-13 14:17 GMT+01:00 : > ah, ok :) > > -- > cinap > >
Re: [9fans] Doesn't 'close' call finish pending 'read' on the same 'fd'?
> another way to implement these kinds of timeouts is > alarm(), see sleep(2). > Yes, I know. But it was just a simplified example to explain my real need: - 'receiver' process - another 'sender' process, which should have a possibility to close the TCP connection, and relase 'receiver' from read() call... Pavel
Re: [9fans] Doesn't 'close' call finish pending 'read' on the same 'fd'?
Hi cinap, hangup(cfd);/* hangup connection, fd stays valid but i/o will error */ > thank you for your explanation and help. Going to use 'hangup' function. ;) Pavel
[9fans] Doesn't 'close' call finish pending 'read' on the same 'fd'?
Hi all,
I am trying such simple program:
#include
#include
static void
closer(int dfd) {
print("closer(%d): 1\n", getpid());
sleep(10 * 1000);
print("closer(%d): 2\n", getpid());
close(dfd);
print("closer(%d): 3\n", getpid());
}
void
main(int, char**)
{
int dfd;
char ddir[NETPATHLEN];
char buf[1024];
long nr;
print("main (%d): 1\n", getpid());
dfd = dial("tcp!192.168.8.61!23", nil, ddir, nil);
if (dfd < 0)
sysfatal("dial: %r");
print("main (%d): 2\n", getpid());
switch (rfork(RFPROC|RFREND|RFMEM)) {
case -1:
sysfatal("fork: %r");
case 0:
closer(dfd);
break;
default:
for (;;) {
print("main (%d): 3\n", getpid());
nr = read(dfd, buf, sizeof buf);
print("main (%d): 4\n", getpid());
if (nr <= 0)
break;
}
print("main (%d): 5\n", getpid());
close(dfd);
break;
}
exits(nil);
}
I want to have one process waiting in 'read' function on opened TCP
connection.
I want to close such a TCP connection from another process (created with
shared file descriptor table).
I would expect that 'read' function is finished with -1...
...but it is not.
Here is the output of the program:
main (2): 1
main (2): 2
closer(3): 1
main (2): 3
main (2): 4
main (2): 3
closer(3): 2
closer(3): 3
You can see that 'main' process is still pending in 'read' function on
already closed 'fd'.
Do I really have to use notes to release the pending 'main' in 'read' call?
Pavel
Re: [9fans] Failed 'reject' function?
Hi Sergey, It seems, you do it incorrectly :) dial(2) has an exact example if using > announce/listen/accept functions. > I do exactly what I need. ;) The sequence announce/listen/accept is very well known to me. But I need both possibilities: - accept incoming connection - reject incoming connection, which failed. Thank to Charles, I finally understood. Pavel
[9fans] Failed 'reject' function?
Hi all,
I tried very simple program on native Plan9:
#include
#include
void
main(int, char**)
{
int afd, lfd;
char adir[NETPATHLEN], ldir[NETPATHLEN];
afd = announce("tcp!*!20540", adir);
if (afd < 0)
sysfatal("listen: %r");
lfd = listen(adir, ldir);
if (lfd < 0)
sysfatal("listen: %r");
if (reject(lfd, ldir, "unauthorized") < 0)
sysfatal("reject: %r");
close(lfd);
close(afd);
exits(nil);
}
I dicovered that the funtion 'reject' fails:
term% 8.out
reject: unknown control request
In 'ratrace' listing I can really see the 'reject' function failure:
348 8.out Open 33a7 0xdfffe9bc/"/net/cs" 0x2 = 3 "" 1412926762174586973
1412926762174841785
348 8.out Pwrite 5577 3 0x7880/"tcp!*!20540" 11 -1 = 11 ""
1412926762186976342 1412926762187086984
348 8.out Seek 3391 0x7ab8 3 0x0 0 = 0 "" 1412926762198637769
1412926762198646989
348 8.out Pread 557f 3 0xdfffe9bc 255 -1 0xdfffe9bc/"/net/tcp/clone.20540"
255 -1 = 20 "" 1412926762200546351 1412926762200619275
348 8.out Close 1225 3 = 0 "" 1412926762221784602 1412926762221865069
348 8.out Open 33a7 0xdfffebe0/"/net/tcp/clone" 0x2 = 3 ""
1412926762223717492 1412926762223849089
348 8.out Pread 557f 3 0xdfffede9 246 -1 0xdfffede9/"4" 246 -1 = 1 ""
1412926762235853725 1412926762235866298
348 8.out Pwrite 5577 3 0xdfffece0/"announce.20540" 14 -1 = 14 ""
1412926762247454802 1412926762247479948
348 8.out Open 33a7 0xdfffede0/"/net/tcp/4/listen" 0x2 = 4 ""
1412926762259028218 1412926771821372908
348 8.out Pread 557f 4 0xdfffede9 246 -1 0xdfffede9/"6" 246 -1 = 1 ""
1412926771844967723 1412926771844984487
348 8.out Pwrite 5577 4 0xdfffedec/"reject.6.unauthorized" 21 -1 = -1
unknown control request 1412926771854330421 1412926771854347185
348 8.out Errstr 5587 0xdfffe914 128"" 128 = 0 1412926771871626615
1412926771871632483
348 8.out Errstr 5587 0xdfffe914 128"" 128 = 0 unknown control request
1412926771887234674 1412926771887239703
348 8.out Pwrite 5577 2 0xdfffe980/"reject:.unknown.control.request." 32
-1reject: unknown control request
= 32 "" 1412926771918100040 1412926771932457957
348 8.out Open 33a7 0x7538/"#c/pid" 0x0 = 5 "" 1412926771947976329
1412926771948085295
348 8.out Pread 557f 5 0xdfffeaa0 20 -1 0xdfffeaa0/"348." 20 -1 =
12 "" 1412926771963591547 1412926771963607473
348 8.out Close 1225 5 = 0 "" 1412926771979115786 1412926771979128359
348 8.out Exits 122d 0xdfffeadc/"reject: unknown control request"
Do you have any explanation?
Did I make some bug in the program?
Thanks in advance.
Pavel
Re: [9fans] Mysterious auth again...?
Hi Stevie, you were right, I missed one point during the server installation/configuration: I forgot 'auth/changeuser bootes'... :( Now 'auth/debug' works well, sorry for a noise. And thank you. Pavel 2014-08-05 11:09 GMT+02:00 Pavel Klinkovský : > Hi Stevie, > > I think this thread on 9fans might help: >> >> http://marc.info/?l=9fans&m=116732560810918&w=2 > > > thanks, going to read it too. > > > >> Read the whole thread, but I think my linked message has the answer. >> Do you start keyfs before listen in your cpurc? >> > > Yes, it was 1st thing I checked... > > Pavel > >
Re: [9fans] Mysterious auth again...?
Hi Stevie, I think this thread on 9fans might help: > > http://marc.info/?l=9fans&m=116732560810918&w=2 thanks, going to read it too. > Read the whole thread, but I think my linked message has the answer. > Do you start keyfs before listen in your cpurc? > Yes, it was 1st thing I checked... Pavel
Re: [9fans] Mysterious auth again...?
Hi Stevie, I know, auth can be tricky... In fact. > Without having further information about your setup I > can only recommend reading this doc: > http://kamalatta.ddnss.de/config/Plan9Tutorial.txt. > Thanks for the link, going to recheck my configuration. > P.S.: I don't know the full background, but connecting as the > hostowner seems to have other prerequisites. > That's what is surprising me. I can 'cpu' the machine with 'bootes' (hostowner) uid. I can 'srv' the machine... I must have something strange there... Pavel
[9fans] Mysterious auth again...?
Hi all,
I am fighting with configuration of '9pccpuf' server.
I have configured user 'bootes' as a hostowner.
I have 'listen', 'keyfs' running.
I can 'cpu' such server from another Plan9 terminal ('9pcf') as user
'bootes'.
I can 'srv' such server from another Plan9 terminal ('9pcf') as user
'bootes'.
BUT:
server# auth/debug
p9sk1 key: dom=xxx proto=p9sk1 user=bootes !hex? !password?
successfully dialed auth server
password for bootesxxx [hit enter to skip test]:
cannot decrypt ticket1 from auth server (bad t.num=0x...)
auth server and you do not agree on key for bootesxxx
I found that it was already solved in the past...
http://comments.gmane.org/gmane.os.plan9.general/55049
but it looks I have a correct sequence in /rc/bin/cpurc, /cfg/server/cpurc,
/cfg/server/cpustart...
Any idea what do I have wrong?
Thanks in advance.
Pavel
P.S.: I cannot make 'auth' system to work with another added user... :(
Re: [9fans] Mistake in Plan9 Mercual port?
> > 1. Is is a new 'python' interpretter implementation? What is a difference > from 'bichued/python’? > > > It’s a ‘new’ CPython 2.7.x version. No dependencies on 3rd party libs > (e.g., openssl, bzip2, …). > Ah, excellent. Going to try... (Personally, I never liked the dependency on openssl...) > 3. I see a little chicken-egg problem, to clone such 'python' version, I > need 'bichued/python' and 'bichued/hg' already installed... ;) > > > No problem there, you can download source tar+gz|bz2 directly from both > bitbucket and selenic. > OK > To build just unpack: > > src/ >cpython/ >hg/ > > and follow the cpython/mkfile rules. > Going to follow your recommendation. Thank you and Erik for your answers. Pavel
Re: [9fans] Mistake in Plan9 Mercual port?
> > https://bitbucket.org/jas/cpython > > you can pull hg directly from the mainline. > Well, I am a little bit confused... 1. Is is a new 'python' interpretter implementation? What is a difference from 'bichued/python'? 2. Is my original problem with mercurial caused by 'hg' or 'python' on Plan9? 3. I see a little chicken-egg problem, to clone such 'python' version, I need 'bichued/python' and 'bichued/hg' already installed... ;) Pavel
Re: [9fans] Mistake in Plan9 Mercual port?
which version of hg are you using? > ...from contrib - bichued/hg > i'm pretty sure that jeff's version use ape/psh to execute > commands, but not positive. it must use psh to be posix-y. > Ah, where is it available? Pavel
Re: [9fans] Mistake in Plan9 Mercual port?
It seems I do not fully understand the discussion about ssh... When I manually make this command (what 'hg' does)... ssh xxx.yyy.zzz "hg init test" ...the result is: bash: hg: Command not found... When I manually modify it... ssh xxx.yyy.zzz 'hg init test' ...the result is... !Adding key: proto=pass server=xxx.yyy.zzz service=ssh user=pavel password: ! ...and the remote repository is successfully created. Therefore I though the problem is wrong usage of " instead of ' in the mercurial port... Pavel
Re: [9fans] Mistake in Plan9 Mercual port?
Hi Steven, For the most part, using HTTP/S repositories will give you the best bang > for the proverbial buck. I see. In fact I tried to create and use Mercurial repository via 'ssh' and 'ftp' (via ftpfs) and none of them works. Pavel
[9fans] Mistake in Plan9 Mercual port?
Hi all, I am using native Plan9. I tried to create a remote Mercurial repository via 'ssh': term% hg init ssh://xxx.yyy.zzz/test --verbose running ssh xxx.yyy.zzz "hg init test" running ssh xxx.yyy.zzz "hg -R test serve --stdio" remote: abort: Bad file number abort: no suitable response from remote hg! Should not it be? running ssh xxx.yyy.zzz *'*hg init test*'* running ssh xxx.yyy.zzz *'*hg -R test serve --stdio*'* Pavel
Re: [9fans] Maximal number of processes
I discovered another interesting feature in my tests: If 'fork' exceeds nproc value, kernel panics. If 'fork' reaches the available memory limit, it is blocked until some memory is released (e.g. by the finish of some process). Pavel 2014/1/10 erik quanstrom > > launching 32000 processes was not possible. the kernel got stuck. > > sloppy statement. it's not clear if the kernel was really stuck or just > hit something exponential. > > > here's one thing that's not immediately obvious, even when running the > > kernel. conv.nmach must be less than 0x7fff/(1000*1000) = 2147 > > to prevent reprioritize from doing silly things: > > > > /* > >* fairshare = 1.000 * conf.nproc * 1.000/load, > >* except the decimal point is moved three places > >* on both load and fairshare. > >*/ > > fairshare = (conf.nmach*1000*1000)/load; > > and, look. the comment's wrong. never read the comments! > > - erik > > <>
Re: [9fans] Maximal number of processes
Hi Steven, conf.nproc = 100 + ((conf.npage*BY2PG)/MB)*5; > if(cpuserver) > conf.nproc *= 3; > if(conf.nproc > 2000) > conf.nproc = 2000; > > In general, you will find that 2000 is the highest allowable due to > limits imposed by proc.c. but if I understand it correctly it is just "soft limit", not a "hard" one inborn in CPU architecture. By the hard limit I consider something like "maximal capacity of GDT, LDT" or something similar, if exists. > At the end of the day, the only way to be sure is to read the source. > I did (of course, I am not an expert on Plan9 kernel), and I did not find any "hard limit" there. However to be sure I issued my question here. Pavel
[9fans] Maximal number of processes
Hi all, I would like to know whether there is any hard (based on CPU architecture) limit of maximal number of processes in Plan9 on Intel or ARM. I do not think the soft limit like the lack of memory... ;) Thanks in advance for any hint. Pavel
