-={ 2021-04-08 16:35:00.799903009+00:00 }=-

Hey Benny!

gcc-10.3.0 has just been released.

Life is good,
Maurice

... Freond deah feor ge neah; byð near nyttra.
A friend is useful, far or near; the nearer the better.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Re: shiney new
By: Maurice Kinal to Benny Pedersen on Thu Apr 08 2021 04:35 pm

gcc-10.3.0 has just been released.

Pat was speaking of this yesterday in slackware's what's new. I'm expecting an update soonish an then a beta of slackware 15.0. The release is approaching. :)

Ttyl :-),
Al

... I'm sure it's all clearly explained in the Zmodem DOC's
--- SBBSecho 3.14-Linux
* Origin: The Rusty MailBox - Penticton, BC Canada (1:153/757.2)

-={ 2021-04-08 21:27:11.428324424+00:00 }=-

Hey Alan!

gcc-10.3.0 has just been released.

Pat was speaking of this yesterday in slackware's what's new.
I'm expecting an update soonish an then a beta of slackware 15.0.

I am calling it 15.0-ish. I have it installed on a laptop and it works rather well but I haven't really done much with it. No fidonet on there ... yet. My wireless networks work good though and am working on a 5GHz host which is where the laptop will connect to as it's default connection. So far so good.

I plan to start a new build of x86_64-motorshed-linux-gnu very soon. My last attempt with building a toolchain using the release candidate failed on coreutils which is a mission critical package. I'll know shortly if the problem still exists and may have to hold off on an install until I discover a fix.

Life is good,
Maurice

... Feorcyþðe beoð selran gesohte þæm þe him selfa deah.
Distant lands are best sought by one who is worthy in himself.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Hello Maurice!

08 Apr 2021 16:35, Maurice Kinal wrote to Benny Pedersen:

gcc-10.3.0 has just been released.

----- gcc begins -----
[I] sys-devel/gcc
Available versions:
(6.5.0) 6.5.0-r3^t
(7.5.0) 7.5.0-r1^t
(8.4.0) 8.4.0-r1^t 8.4.0-r2^t
(9.3.0) 9.3.0-r1^t 9.3.0-r2^t
(10) 10.2.0-r5^t ~10.3.0^t **10.3.1_pre9999*l^t
(11) **11.0.1_pre9999*l^t
{ada altivec awt cilk custom-cflags +cxx d debug doc fixed-point +fortran gcj go graphite hardened jit libssp lto mpx multilib +nls +nptl objc objc++ objc-gc +openmp +pch pgo +pie +sanitize +ssp systemtap test valgrind vanilla (+)vtv zstd}
Installed versions: 10.2.0-r5(10)^t(11:44:40 PM 03/08/2021)(cxx nptl openmp pch pie sanitize ssp -ada -d -debug -doc -fixed-point -fortran -go -graphite -hardened -jit -libssp -lto -multilib -nls -objc -objc++ -objc-gc -pgo -systemtap -test -vanilla -vtv -zstd)
Homepage: https://gcc.gnu.org/
Description: The GNU Compiler Collection

----- gcc ends -----

not yet in gentoo

i like to learn more on how to disabled compressed memory

symtoms is when doing

swapoff -a

then in top there is still threads that show it have or still have usage on swap memory, is it sign of memory leak ?


Regards Benny

... too late to die young :)

--- Msged/LNX 6.1.2 (Linux/5.10.27-gentoo-dist (x86_64))
* Origin: I will always keep a PC running CPM 3.0 (2:230/0)

-={ 2021-04-14 21:57:54.332576812+00:00 }=-

Hey Benny!

not yet in gentoo

I see it is now in Slackware-15.0-ish.

i like to learn more on how to disabled compressed memory

Compression is evil. I would definetly get rid of that.

swapoff -a

I haven't used swap in around 25 years now. What for? DDR4 is cheap, or at least was the last time I checked and infinetly better than swap. Why would you do this to any machine?

is it sign of memory leak ?

I don't know. All I know is that real memory seems to be working fine and thus no need for swap to act as drag on any system.

Wait a minute! Yes I have used swap lately on a raspi but that was only because it lacked real memory and I needed more to enable 64-bit gcc and friends (aarch64-unknown-linux-gnu). However that fell by the wayside once I was successful getting it to work. Pure 64-bit too. A real shame that turned out to be although I did learn much from the experience, the biggest lesson being raspi's require more real memory. 8G would have been nice. Oh well ... live and learn eh?

Life is good,
Maurice

... Geara is hwær aræd.
A resolute man is always prepared.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Hi Maurice,

On 2021-04-14 21:57:54, you wrote to Benny Pedersen:

swapoff -a

I haven't used swap in around 25 years now. What for? DDR4 is cheap, or at least was the last time I checked and infinetly better than swap. Why would you do this to any machine?

On a busy server a little bit of swap can be usefull, so the kernel doesn't start killing random processes right away when it runs out of real memory. This gives the administrator some time to take messures, without risking unplanned interrupted services...

Bye, Wilfred.

--- FMail-lnx64 2.1.0.18-B20170815
* Origin: FMail development HQ (2:280/464)

-={ 2021-04-15 22:40:47.196178230+00:00 }=-

Hey Wilfred!

On a busy server a little bit of swap can be usefull, so the
kernel doesn't start killing random processes right away when
it runs out of real memory.

After +20 some odd years I have yet to see that happen.

The one instance where swap was needed was on the raspi when attempting to update to aarch64-unknown-linux-gnu and once that was done I turned off swap and then later eliminated it from the microSD card once I was satisfied it wasn't needed anymore. Swap on SD cards isn't a good idea to say the least but I was willing to sacrifice one for the cause.

The card still works but I only use it for storage these days as I retired the raspi once I was satisfied I did all I could for that cause and that other than a learning experience there is nothing there of any real value to me personally, especially the crippled onboard wireless chip. What a waste.

Life is good,
Maurice

... Mon sceal... gebidan þæs he gebædan ne mæg.
One must wait for what cannot be hastened.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Hello Maurice!

15 Apr 21 22:40, Maurice Kinal wrote to Wilfred van Velzen:

On a busy server a little bit of swap can be usefull, so the
kernel doesn't start killing random processes right away when
it runs out of real memory.

After +20 some odd years I have yet to see that happen.

Just start an application that runs into a memory leak...


Regards,
Gerrit

... 8:44AM up 177 days, 18:45, 8 users, load averages: 0.75, 0.58, 0.50
--- Msged/BSD 6.1.2
* Origin: We are a nation of innovations (2:240/12)

-={ 2021-04-16 08:47:23.588700205+00:00 }=-

Hey Gerrit!

After +20 some odd years I have yet to see that happen.

Just start an application that runs into a memory leak...

Offhand I am unsure what my motivation would be but just for fun what application do you recommend?

Life is good,
Maurice

... Sprec ofter embe oðres monnes weldæde þonne emb ðine agna.
Speak more often about other people's good deeds than about your own.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Hi Maurice,

On 2021-04-15 22:40:47, you wrote to me:

On a busy server a little bit of swap can be usefull, so the
kernel doesn't start killing random processes right away when
it runs out of real memory.

After +20 some odd years I have yet to see that happen.

I was talking in general, not your particular case...

Bye, Wilfred.

--- FMail-lnx64 2.1.0.18-B20170815
* Origin: FMail development HQ (2:280/464)

Hello Maurice!

16 Apr 21 08:47, Maurice Kinal wrote to Gerrit Kuehn:

Just start an application that runs into a memory leak...

Offhand I am unsure what my motivation would be but just for fun what application do you recommend?

As memory leaks are usually bugs, there is nothing I can specifically recommend here. I can only say that I experience this kind of issue every once in a while, and without swap a system can run out of memory pretty quickly. Sometimes the kernel will start terminating processes at random before you can even think CTRL-C.


Regards,
Gerrit

... 1:26PM up 177 days, 23:27, 8 users, load averages: 0.10, 0.22, 0.21
--- Msged/BSD 6.1.2
* Origin: So come and try to tell me (2:240/12)

-={ 2021-04-16 12:33:14.827378778+00:00 }=-

Hey Wilfred!

I was talking in general, not your particular case...

That makes me even more curious. In general, what server apps should I be on the look out for?

Life is good,
Maurice

... Se ðe him ealne weg ondræt, se bið swylce he sy ealne weg cwellende.
He who is always afraid is like one who is always dying.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's EuroPoint - Ladysmith BC, Canada (2:280/464.113)

Hi Maurice,

On 2021-04-16 12:33:14, you wrote to me:

I was talking in general, not your particular case...

That makes me even more curious. In general, what server apps should I be on the look out for?

Who's talking about apps? We were talking about swap being a good idea on busy servers...

Bye, Wilfred.

--- FMail-lnx64 2.1.0.18-B20170815
* Origin: FMail development HQ (2:280/464)

-={ 2021-04-16 13:31:05.952678587+00:00 }=-

Hey Gerrit!

As memory leaks are usually bugs, there is nothing I can
specifically recommend here.

Just as well since I probably avoid running whatever it is anyhow. So far everything has been behaving itself and there is definetly more than enough ram, overkill I'd say.

I can only say that I experience this kind of issue every once
in a while, and without swap a system can run out of memory
pretty quickly.

Understood. However the two mini-servers currently being used here only have an SSD which apparently is an issue for swap. For any serious file storage I have a server on the private LAN that takes care of regular sata drives as it can hotswap them if and when it matters. I suppose if push comes to shove I could supply swap over the network but offhand I don't like this idea.

Sometimes the kernel will start terminating processes at random
before you can even think CTRL-C.

No keyboard(s) on them. Occasionaly a monitor to watch a new system boot up but that doesn't happen too often.

On the one I am writing this reply on, I am using a ssh login from my main workstation which is also on the private LAN. Both the Brain and the EuroPoint are running on wireless hotspots but there is no access from the wireless interface. That is just for the neighbours who need them to access the internet with their smart devices, mostly smartphones. So far the lack of swap hasn't been a concern and probably has saved money given my frugal nature of doing writes to their SSDs. +20 years ago it was a bootable CompactFlash that kickstarted this whole idea for me. That was around 1999-ish. No swap then either.

Life is good,
Maurice

... Gold geriseþ on guman sweorde... sinc on cwene.
Gold is fitting for a man's sword, precious things for a woman.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

-={ 2021-04-16 13:59:58.450659491+00:00 }=-

Hey Wilfred!

We were talking about swap being a good idea on busy servers...

I'll have to take your word on it.

Life is good,
Maurice

... Scyle monna gehwylc mid gemete healdan a wiþ leofne ond wið laþne.
Every man should act with moderation both to friends and foes.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Hello Maurice!

15 Apr 21, Maurice Kinal wrote to Wilfred van Velzen:

done I turned off swap and then later eliminated it from the microSD
card once I was satisfied it wasn't needed anymore. Swap on SD cards isn't a good idea to say the least but I was willing to sacrifice one
for the cause.

The card still works but I only use it for storage these days as I
retired the raspi once I was satisfied I did all I could for that
cause and that other than a learning experience there is nothing there
of any real value to me personally, especially the crippled onboard wireless chip. What a waste.

I'd like to lol a little bit. ;-) The raspi is designed as a cheap learning plattform for everyone. Your lession learned was that swap can be useful. The next lesson would be that it's better to avoid swap for performance but this one you knew already. People new to IT would learn that the hardware resources need to match the task(s). Fullfilled its task the raspi had.

Regards

Kai

--- GoldED+/LNX 1.1.4.7
* Origin: Monobox (2:240/77)

-={ 2021-04-16 17:16:15.981413948+00:00 }=-

Hey Kai!

I'd like to lol a little bit. ;-)

Me too. It beats the alternative by a long shot.

Fullfilled its task the raspi had.

For sure. May it rest in peace.

Life is good,
Maurice

... Þæt folc bið gesælig... and gesundful þurh gesceadwisne reccend.
A people is made happy and prosperous by a wise ruler.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

-={ 2021-04-16 19:06:29.747324022+00:00 }=-

Hey Maurice!

Ooooops!

Life is good,
Maurice

... Swa cystigran hiwan, swa cynnigran gystas.
The more generous the household, the more noble the guests.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's EuroPoint - Ladysmith BC, Canada (2:280/464.113)

-={ 2021-04-16 19:23:10.281112415+00:00 }=-

Hey Maurice!

Memory leak? :::snicker:::

It should be fixed now but we'll soon see.

Life is good,
Maurice

... Se anweald næfre ne biþ god, buton se god sie þe hine hæbbe.
Power is never good, unless the one who possesses it is good.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Hello Maurice!

16 Apr 2021 19:23, Maurice Kinal wrote to Maurice Kinal:

-={ 2021-04-16 19:23:10.281112415+00:00 }=-

Hey Maurice!

Memory leak? :::snicker:::

i have msg leak

reply i write vanish before stored and sent :(

yes i need new ebuilds for the husky cvs builds, hopped that sf.net was not make the cvs read only

It should be fixed now but we'll soon see.

else there will be free ice cream ?


Regards Benny

... too late to die young :)

--- Msged/LNX 6.1.2 (Linux/5.10.27-gentoo-dist (x86_64))
* Origin: I will always keep a PC running CPM 3.0 (2:230/0)

-={ 2021-04-19 14:46:08.179073139+00:00 }=-

Hey Benny!

i have msg leak

Mine turned out to be a msgid leak. I am surprised it got anywhere.

reply i write vanish before stored and sent :(

I hate when that happens.

It should be fixed now but we'll soon see.

else there will be free ice cream ?

Seeing it is fixed then yes.

Life is good,
Maurice

... Maþþum oþres weorð, gold mon sceal gifan.
One treasure deserves another; gold should be given away.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Wilfred van Velzen wrote to Maurice Kinal <=-

@MSGID: <607806F8.314.fidonet_linux@caughtinadream.com>
@REPLY: <60776C55.312.fidonet_linux@caughtinadream.com>
@TZ: 0078
Hi Maurice,

On 2021-04-14 21:57:54, you wrote to Benny Pedersen:

swapoff -a

I haven't used swap in around 25 years now. What for? DDR4 is cheap, or at least was the last time I checked and infinetly better than swap. Why would you do this to any machine?

On a busy server a little bit of swap can be usefull, so the kernel doesn't start killing random processes right away when it runs out
of real memory. This gives the administrator some time to take
messures, without risking unplanned interrupted services...

Always leave some swap on a Linux system even if you have 32, 64GB, or
even a TB of memory. I've seen applications completely consume all
available memory and, thankfully memory was monitored, was able to kill
the processes that consumed all the memory to prevent the server from
going down. Ultimately, it's a safety net, it is stupid not to use it.

This is similar to partitioning out the file systems on a Linux
server. Keeping everything in one file system will bite you in the butt
if you have a directory fill up.


Brian Klauss <-> Dream Master
Caught in a Dream | caughtinadream.com a Synchronet BBS

... Gone crazy, be back later, please leave message.
=== MultiMail/Mac v0.52
--- SBBSecho 3.14-Linux
* Origin: Caught in a Dream - caughtinadream.com (1:104/116)

Hello, Gerrit Kuehn.
On 4/16/21 8:44 AM you wrote:

After +20 some odd years I have yet to see that happen.

Just start an application that runs into a memory leak...

--------cut memleak.c -------
#define gigs = 8 // your gigabytes here
#include <stdlib.h>
int main() {
long int i=0;
while(i<1024*1024*1024*gigs/2) {
void *nofree = malloc(++i);
}
return 0;
}
---------cut-------

Try this w/ gcc -o memleak memleak.c; ./memleak

No free() after malloc() should give memory leak, isn't it?

--
Best regards!
Posted using Hotdoged on Android
--- Hotdoged/2.13.5/Android
* Origin: Android device, Milky Way (2:5020/2141.808)

Hello Maurice!

19 Apr 2021 14:46, Maurice Kinal wrote to Benny Pedersen:

i have msg leak

Mine turned out to be a msgid leak. I am surprised it got anywhere.

lets hope husky will be debugged with gcc 11

reply i write vanish before stored and sent :(

I hate when that happens.

#metoo

It should be fixed now but we'll soon see.

else there will be free ice cream ?

Seeing it is fixed then yes.

my husky ebuilds works if sf.net had not turned cvs into read-only

still have to change ebuilds to be using github where husky now lives


Regards Benny

... too late to die young :)

--- Msged/LNX 6.1.2 (Linux/5.10.27-gentoo-dist (x86_64))
* Origin: I will always keep a PC running CPM 3.0 (2:230/0)

Hello Oleg!

25 Apr 2021 16:13, Oleg Artemjev wrote to Gerrit Kuehn:

Try this w/ gcc -o memleak memleak.c; ./memleak

No free() after malloc() should give memory leak, isn't it?

postgresql have problems aswell, imho when issue swapoff then top should not show that its still using swap filesystem for that threads in top, if this happens its imho sign of kernel problem, or really mem leak


Regards Benny

... too late to die young :)

--- Msged/LNX 6.1.2 (Linux/5.10.27-gentoo-dist (x86_64))
* Origin: I will always keep a PC running CPM 3.0 (2:230/0)

Hello Benny!

29 Apr 21 15:35, you wrote to Oleg Artemjev:

postgresql have problems aswell, imho when issue swapoff then top
should not show that its still using swap filesystem for that threads
in top, if this happens its imho sign of kernel problem, or really mem leak

You can take nearly any JAVA app. Java heap usualy exhausted...

Karel

--- GoldED+/LNX 1.1.5-b20180707
* Origin: Plast DATA (2:423/39)

-={ 2021-04-29 18:21:19.722786421+00:00 }=-

Hey Benny!

still have to change ebuilds to be using github where husky now
lives

I haven't done a husky build for well over 15 years now. I did note the switch to github not too long ago. I might try it later to see about any gcc11 issues it might have. So far only rustc seems to have issues with the 'upgrade' although I had to bump llvm to 12.0.0. Overall I am surprised I got as far as I have with it. For now it is only one machine with an extra-special bootable gcc11 partition on a nvme drive.

Life is good,
Maurice

... Don't cry for me I have vi.
--- GNU bash, version 5.1.4(1)-release (x86_64-motorshed-linux-gnu)
* Origin: Little Mikey's Brain - Ladysmith BC, Canada (1:153/7001)

Hello Karel!

29 Apr 2021 20:02, Karel Kral wrote to Benny Pedersen:

You can take nearly any JAVA app. Java heap usualy exhausted...

never needed java, but getjar apps to my nokie e51 worked :)


Regards Benny

... too late to die young :)

--- Msged/LNX 6.1.2 (Linux/5.10.27-gentoo-dist (x86_64))
* Origin: I will always keep a PC running CPM 3.0 (2:230/0)