Is there nothing like mhddfs for FreeBSD?

Basically a way to mount a whole bunch of drives into one big filesystem (at file-level, so not block level like gconcat)

mount how exactly?

Well on Linux I can have a bunch of drives with the same folder names but different drives

With MHDDFS I can make one virtual mount, in which I can see that folder but all the files from the same drives combined

And if I create a new file it will pick the most empty drive

It's really handy to have a NAS with totally different sizes of drives and still see them in one and the same tree

it looks like a fuse filesystem, which should mean it should be portable to freebsd

and lo, there already exists a port

sysutils/fusefs-mhddfs

Hmmm the one on Linux I use is not fuse I think

I don't like fuse so much due to its instability

But ok if it's the only way... Thanks!

what linux distro are you using?

because every mention of mhddfs that I find mentions that it is a fuse filesystem

Debian / Openmediavault

But I'd like to move to FreeBSD for real ZFS support

That does explain a lot because even on debian it's a bit unstable lol

But ok it should work, it's mostly for reading anyway

Thanks!

anything like that that works at file level is always going to be a bit of a hack

Yeah but block level is really a problem if you want to be able to just rip one drive out and replace it with a bigger one

And just copy the files over that were on it

I have 8 drives and I can't upgrade them all at the same time so any RAID solutions including RAIDZ are just out of the question, sadly

Once a year or so I buy a new drive, replace the smallest one in the NAS and the old drive goes into my backup box where I basically use all the old drives as tape cartridges :3

With tar multi-volume which I also have to check on FreeBSD, now that I think of it

There is not yet any support for multi-volume archives.

so sayeth `man tar`

ouch ok

Thanks!

Maybe I can build gnu tar for this or something

Willcheck

gnu tar is available as a port/package

Ahh nice thanx

That will do for this purpose

I know it's a bit of a weird setup all in all but without money to upgrade all my drives, I don't really see a better way

A friend recommended unraid but it's all so proprietary, I don't like it

And I really want block checksumming

hi

Hi people, anyone have experience using ngircd?

|F9|: that name sounds familar; perhaps long ago

I just can't find any information about vHosts.

that's not really a freebsd issue

I accidentally destroyed the zfs dataset containing my home directory. I didn't do anything destructive after that--I immediate exported the pool. The internet indicates that it may be possible to recover it with: zpool import -N -T txg-id -o readonly yroot, where txg-id is a txg from before I destroyed the dataset

the txg which removed it was 336735. If I use any id above 336735, the import succeeds (but of course lacks my data). If I use 336735 or 336734, however, it hangs for a while and then prints: cannot import 'yroot': one or more devices is currently unavailable

rather confused by this. Any suggestions? Is there anything I can do?

moon-child: restore from backup and not make that mistake again.

" immediate exported the pool" you mean zpool export -f ?

I don't remember

i don't expect you to get any further data back.

you might be able to use zdb to display the transactions

why is it printing that error? Presumably the device is available. (Wasn't in raid or anything like that, so there's only one device. Perhaps it's flaking out? Seems unlikely, as I got it pretty recently, but maybe it would be helpful to copy everything onto another disc?)

it's printing the error because it can't import that transaction

rtprio: Yeah I know, sorry, just nobody is on #ngircd, and I have no idea where else I can ask.

seems like a strange thing to print in that situation :P but ok. What am I looking for in zdb output?

you're going to have to do some research first as it could be more destructive than what you did

it's like, if you're using zdb, you're already way off the rails of 99% of zfs usage

|F9|: vhosts for the client or... server ?

rtprio: for clients, eg cloaks.

I have not found literally anything, will search more of course.

if it's not in the documentation, perhaps it's not supported

Title: ngircd/sample-ngircd.conf.tmpl at master · ngircd/ngircd · GitHub

rtprio: Features .. > Supports "cloaking" of users

ah I see

My wpa_supplicant.conf only contains the ctrl_interface='' option, I start wpa_supplicant -B and try to scan with wpa_cli which gives me <3>CTRL-EVENT-SCAN-RESULTS\n<3>CTRL-EVENT-NETWORK-NOT-FOUND

I dont have WPA setup on the interface in rc.conf so wpa_supplicant doesnt auto start

How should one interpret the error messages?

Hi, It's fine that nginx logs are rotated automatically but I would like to know, what created file /etc/newsyslog.conf.d/local.conf ...Does anyone know?

I have this file on two servers and on third is missing, so probably I created it myself and don't remember it. So i take my question back, mystery is solved :)

vyborne!

Someone know if there is a way to have a bigger / wide-screen resolution with GOP/bhyve?

pvalenta: pkg which /etc/newsyslog.conf.d/local.conf say anything?

meena0, actually no,... I had create it myself year ago and forgot it. I am sorry for stupid question

pvalenta: brains are like that sometimes

hi, i'm having troubles to send system reports via smtpauth ( too complicated with bsd built-in sendmail )

is there a way to change mail sending mechanism that way to send reports by using, say other programs ?

or let say substitute it via python script

I don't know how to parse that question.

:)

i need a simpler way to send e-mail reports ( in general ) by using smtps / smtpauth

Do you mean the reports generated by periodic(8)?

periodic, smartd, .....

smartd isn't part of the base system.

but still sending mails using /usr/bin/mail

So find something that replaces /usr/bin/mail

dma does, as it's already replaced sendmail in 14-CURRENT.

uh, really ????

do you have a link to this awesome news ?

nerozero: cgit.freebsd.org/src/commit/?id=a67…25ff3e58b072a60b633e442ee1d33e47f7f

Title: src - FreeBSD source tree

too bad that wasnt opensmtpd instead \o/

dkeav, thanks

CmdLnKid: for a long time people were asking for dma, and it couldn't be imported to base because the cluster admin team practices dog-fooding - and when dma finally got imported, because I got the features needed, you want a replacement that nobody knows if it does the right thing?

Also, OpenBSD doesn't exactly have a great trackrecord of accepting patches, cf. dhclient in FreeBSD ending up as its own fork because they refused to accept the capsicum patches and a bunch of other things.

whelp at least the syntax matches up with openbsd pf and can easily be configured in 5 lines of a config but i mean that wouldn't make sense or anything. but surprisingly since my involvement in 2001 it hasn't all be just moved to ports since its consistently been bikeshedded

but hey what would anyone else know, i haven't heard a damn thing about dma before ... $today

dma is great ...I use it on all freebsd instances except mail server where I use opensmtpd

how do I monitor the data flow between my machines?

I think I want to try and lower amount of data flowing, like make apps ask for smaller amount of data to get job done

/29/29

meena could you paste me that review link again, please

tsoome: reviews.freebsd.org/D40094

Title: ⚙ D40094 Add a virtio-gpu 2D driver

ugh, how do you mark individual comment in bugzilla as spam? (bugs.freebsd.org/bugzilla/show_bug.cgi?id=251674#c13)

Title: 251674 – libc++: std::wcout does not use global locale set via setlocale()

meena thanks

yuripv: usually you edit and replace the contents with SPAM

Hello people, Please forgive me, if what i ask sounds rude or lazy. I wanted to move away from Linux and so after 3-4 months of reading and comparing, i finally decided to move to free-bsd. This was a month ago. Yes I am that useless. But once i installed freebsd, my laptop (Thinkpad E450) gets heated even while the systemm is in bootloader process. I tried finding a reason, but couldn't. I

finally gave up and I am currenlty in limbo, using a temporary linux distro to just work with. I have only one machine. I would love to move to freebsd and want to give one more try.

Has anyone been through this before? is it how this is? that, the freebsd kernel works differently and sso this is expected? or is ther something i can do about it and finally move to freebsd, if so please direct me how.

thanks in advance

I am a fairly advanced user and have been using linux for the past 10 years, the free bsd tools are different, and i am willing to learn, but the heat issue scares me, as i have only one machine, my Thinkpad E450 and i can't afford one if that's gone.

I've mostly used gentoo and arch (70%) and have recently swotched to void before taking a plunge to freebsd

anything else please ask

sozuba: most Systems don't initialize why power saving infrastructure at that stage

meena0: forgive my ignorance, but this doesn't happen in linux.

After going through about bsds, especially freebsd i thought i found my place, but this heating issue is a huge pain. If this can be sorted, i could start from there and work my way in learning the different tools

sozuba: Are the fans being driven? Did you enable frequency controls?

i also plann to post in the forum, and wait for a while to see if something good comes out for me, if not i have no other option but to go back :(

sozuba: I find that FreeBSD tends to lag with power management, as you're seeing, but you should be able to keep it from running hot.

mason: at that time, yes i checked, i could here the fans running and i have no idea about the frequency controls. May be these are something i could start with. Can you post me a link to these resources so i could read them?

please

sozuba: what stage of boot exactly are you concerned about?

sozuba: man.freebsd.org/cgi/man.cgi?query=powerd&sektion=8&format=html for one

Title: powerd(8)

mason: that's a relief to hear. If the running hot issue can be solved, i would be greatful

sozuba: Also, there are folks with out of tree projects that will handly SMAPI for you.

handle*

RhodiumToad: the issue happens right after powering on, as soon as the first bsssd bootloader texts come in. INitially i thought it was becuse i had encrypted my disk, but even without it, this happend.

mason: i will definitely read through the link. thanks

sozuba: at that point the bootloader (there is no kernel yet) is typically using 100% of one cpu core, and this will typically be true until you enter the kernel

mason: tree projects? freebsdfoundation.org/project/flattened-device-tree-project <- this?

Title: Flattened Device Tree Project | FreeBSD Foundation

sozuba: Was thinking of github.com/sghctoma/thinkpad-smapi - it reads but doesn't write correctly on my T420 but I bet it's a minor correction either way

Title: GitHub - sghctoma/thinkpad-smapi: Thinkpad SMAPI driver for FreeBSD

"out of tree" as in not maintained as part of FreeBSD

sozuba: the question is, how long are you leaving it in this state and why is it getting noticably hot?

things like powerd only apply after you're in multiuser, at which point the problem is likely gone anyway

RhodiumToad: oh okay, but even after I've logged in the system is hot, after a week of this i got scared and didn't want to destory the only and last option i have for a pc

RhodiumToad: these are the questions i tried finding an answer for, but in vain.

But all your response has given me some encouragement to go for another try. As has been many cases, i've usually had one or two tries to get fully into something, and i hope it works here too

but very scared to kill my pc.

mason: will definitely have a look

sozuba: I mean, how long between starting to boot and seeing the kernel startup messages (the ones after --<<BOOT>>--)

i would say even before the first text message of starting the kernel or a second after the first line popos up

pops*

that's not what I meant

the time between starting to boot and getting those kernel messages is time spent using 100% of one cpu core. how long is that for you in seconds?

RhodiumToad: May be a 1-2 seconds, thats something I will have to check, as I don;t remember. What i will do is, I will try reinstalling free-bsd and let you know.is that fine? I will remember your user handle and tag you, if that is okay with you.

anyone seen a TUI wifi manager?

if only a few seconds of cpu usage is enough for the machine to become noticably hot, that sounds like a problem

RhodiumToad: I first thought may be my hardware is done, but after re-installing linux, it works without any heat.

is this using UEFI, legacy bios, or something else?

there's so many abbev's one can only assume TUI is a TCL user interface to wifi management

tui = text user interface, e.g. written with curses or similar

if you know tcl and the user commands well enough it isn't hard to form your own

My be, what i will do is, try to have more patience and re-install free-bsd, while having myself connected over here through my mobile or my family member's laptop/pc. I will also post in the forum if needed.

in that case use dialog

RhodiumToad: UEFI

;-)

sozuba: also, what freebsd version?

RhodiumToad: 13.1

any particular reason for that over 13.2 ?

RhodiumToad: nothing particular, my logic was that that was the most stable

.1 versus .2 common now

my experience is that if you want to go for a -release version, there's no reason not to pick the latest one on the major version branch you chose

RhodiumToad: what version do you suggest?

13.2

RhodiumToad: cool, will work on that this time. I will note down the time it takes and stuff. I hope this works

great experience but if you previously ran a .X there is no reason why <.X wouldn't do it

fbsd is that stable

CmdLnKid: that's what i want to experience :)

nit picking bullshit

you are fine on 13.1

"vm-bhyve is using the zfs boot loader from the host machine" lists.freebsd.org/archives/freebsd-current/2023-June/003825.html -- interesting

Title: Re: CURRRENT snapshot won't boot due missing ZFS feature

id find it very rare that you can't do 13.1 vs 13.2 in this day and age

at this point in time, i just want to know what causes the heat on my system, is it to be expected with the way freebsd kernel is, or is it happening only to me. If so, why? and can it be sorted.

that is unless you were trying to do that at an age of 5.2

;)

CmdLnKid: oh, was it that much of a difference then?

it was yes

right now backing up my data before i can start re-install

fbsd 5 was a shitshow

CmdLnKid: ah cool, good to know. HOw long ago was that?

sozuba: it's not the kernel, it's (if anything) the boot programs (loader.efi in this case)

CmdLnKid, truth.

2007 ish ? maybe

possibly before

2003

RhodiumToad: fair enough, if so, is there an alternative loader i can use? like grub?

thanks rhodiumtoad

its been that long

RhodiumToad: wow, that long ago. ALso that's the year i wish i got respwaned. SO much in my life could have changed, :d

sozuba: well first we need to confirm that's what the problem is and that after the kernel is running, things cool down properly

sozuba, hindsight is always 20/20

fbsd 5 was a shitshow but it was a mostly necessary one, I think

CmdLnKid: haha, treu and well said :D

rhodiumtoad, i agree. its what got us here today for sure

RhodiumToad: will do, i will try to confirm.

How much DragonflyBSD do you use?

me? none

... toast

I never liked dillon's code

who cares about the code unless it doesn't work

the idea behind it is what matters and whether its understood

given my previous experience with his code didn't encourage me to try out his freebsd fork, let's put it that way

He uh. . . was pretty famous for commits that broke. . . a lot.

yeah great innovations but never fit in my framework

He had big ideas for the OS, but didn't get buy-in from others.

1'sand commits 1'sand fails to get to the right one

Caused the FreeBSD community a lot of pain, before he finally took his ball and went home.

lol

Took his ball and built his ouwn field.

LIV

Hammer is interesting...

don't shoot me

DMA is quite good and useful.

Erhard, there's like five people that really love his field.

I know, not many. I ran a DragonFly box for a while.

I dabled a bit.

Didn't see much advantage, though.

It just worked.

It was such a small and new community, one could make big splash.

I still use DMA on most machines for basic mail tasks.

as good as DF is and i commend it but still won't run it in production even DMA

I mean, FreeBSD still included senmail by default? WTH?

which astonishes me in a way that it was included in base

no disrespect to the devs

I actually liked having sendmail and bind in the base. >=]

crtxreavr, i may just be old school like that

I switched to POstfix back in 99 maybe .

But even it is too much for most mahcines that just need to send out logs.

Postfix was nice if I needed an actual MTA.

RhodiumToad: No doubt, me too.

tested and time proven i can expect the flaws

i am going to re-loggin from a family memeber's pc that i have borrowed, brb

DMA works so nicely to just send out logs through a smarthost.

But the default local-only mode for the included sendmail. . . was just the right size for most deployments.

last time I checked dma it didn't _quite_ managed to cover my use-case

erhard, so did sendmail

sendmail was way to complex and buggy for that

Remember, I was using it in the 90s... It was known for poor coding and security

buggy

Like why I still won't use BIND

sendmail configuration is still a nightmare

That too.

m4 BS

thats what someone says when they are jetting over a sandhill at the local dune

back

mendsail

The sendmail config sytnax is a bit. . . archaic.

lack of knowledge of m4 is no excuse

But it's a very mature platform, despite a few bumps along the way.

Some people still use qmail. It was last updated in 1999, I think.

RhodiumToad just tagging you in this pc so i remember

lol. Knowledge of m4 these days is not so useful.

:D hope you don;t mind

Patch branches are maintained down to the present day.

Autoconf stilll uses m4 though, I believe.

Been a while

it does, unfortunately

I liked autoconf, other than the m4 crap ;-)

things must evolve in other words tho

even if it causes bs

Remember when perl was in base?

lololol

I still use perl

pepperidge farms remembers CrtxReavr

lxghtnxng, lol

I couldn't help but LOL when perl was replaced with awk.

RhodiumToad "it does unfortunately" was that for me?

replaces ?

sozuba: no

no

It has its problems, but at least it is basically C lexecon, and ir doesn't care about whitespace.

That sounds like a dig against Python.

Absolutely.

I like Python quite a lot.

>s/perl/awk/

Perl never really meshed with my brain.

????

Why not Tcl?

I just wish it used the C lexicon, and ignored whitespace.

sozuba: in about 45 minutes I'll have to go out for a while, but I'll be here until then. most likely - you can ignore the other chitchat

And frankly, it's syntax was a little too free-form.

Perl sucks in many ways, don't get me wrong.

Objects are a joke... but in Python too

But that was never the goal of either of those

all languages suck

True

Except C ;-)

especially C :-)

lol

Perl seems to have ended at v5.x.

I know 6.x is out there, but there's like 5 people that use it.

perl 6 kinda ... went weird

That was its heyday

pffft ima deny that

(And only one of them is named Larry.)

RhodiumToad: Called Raku, now.

i may be wrong but i love perl

You're wrong.

I'm wrong too, then

does it scale

no

but im happily wrong

how would you be wrong

language wars should probably go to -social or somewhere

I can...

pfffft

AWS Lambda with perl ;-)

can't talk shit anymore ?

that would go where perl went

fine... lets reincorp tcl into base

CmdLnKid: at #freebsd-social

RE???

Tcl was in base?

lol

someones checking the cvs commits

CmdLnKid: /join #freebsd-social

TCL is for eggdrops.. Does it do anything else ;-) ?

k back on topic

lol erhard

RhodiumToad thank you. I think i will need at least an hour to back up my stiff. Thank you and i hope to catch you here sometime. I am starting the process right now,  i will keep you posted.

SMTP daemon wars?

lol

dns server wars

bind won long ago.

RhodiumToad, no worries with you leaving in 45 minutes, take your time. I'll ask around for issues and if i see you online or active i will tag you, but will keep you posted

I use PowerDNS locally

sozuba: I'll likely be back in around 2 hours

sozuba: can you please stop booping the frog

RhodiumToad is there a forum link i can tag ot message you?

or*

I don't use the forums, this is the best place to catch me

'_'

RhodiumToad thank you two hours is perfect, hopefully i'll catch up

LxGHTNxNGb booping?

did i do something against the rules?

betty boop and a frog ...

unbound or djb

Betty Boop?

:( i dont understand that, not even sure if you guys are being playful or i am actually breaking rules by "booping"

That was the rpecursor to Archie

haha

frog, i understood you meant Rhodium, but booping, you meant tagging like honking?

The guy who wrote Archie gave the commencement at my sons gradutaion

tagging

sozuba: e.g. this

it's mostly a furryism tbh

is tagging frowned up here? I am sorry if that's the case, I didn't know

I think Erhard just tagged himself as the old guy.

no

yeah, kinda

Since my son is out of uni

but I think the toad has other engagements right now, and excessive tagging in any event is not appreciated

prolly dma

Erhard: Some people still keep the djbdns codebase alive. Bit quihotic, but...

I use it on some machine.

Either patched or the ndjbdns variant

LxGHTNxNG oh okay that's unerstandable. I didn;t want to bother, just that i thoguht he would miss my message in the volley of texts people were having (chitchat)

sorry to tag you too, habbit

no I get it

i always thought if you don;t tag it is rude,because expecting someone to look at your message among a flood of messages is bad

Lots of tagging, a veritable Banksy here.

so i always tag

banksy, damn. Appart from learning about freebsd, i should also learn the lingo here. :d

it makes some sense

Alright, I am outta here. catch you all later.

Banksy is a graffiti artist. Don't fully know where but I think England.

cya soon Erhard

if banksy was a bsd user id be impressed

yup just did a websearch, :)

If he uses a un*x it would be FreeBSD.

Or maybe NetBSD.

the boot loader art would be much more impressive

lol.. OK. Cya

Baksny Beastie

Banksy Beastie even

lol lets hire him

hang on. Where does this all go back to? What all problem does sozuba have that needed RT's (abbreviated to avoid causing a notification sound) attention?

can you give me 15 mins, i will be back and i will explain?

id hate to see freebsd shred itself tho

LxGHTNxNG: laptop running very hot prior to kernel startup

brb

okay

back

LxGHTNxNG as RT mentioned, yes laptop running hot priort to kernel startup - short

I am new to bsd,after some research comparing distros and OS and i finally found freebsd to be perfect for me, and installed it.

right

What model of laptop do you have and what CPU?

This was a month ago. i tried for a week to get the issue sorted.But i couldn't i lost hope and installed a temp linux distro to carry on my work. But i wanted to give it one more go, to see if somethign can be done

I use Thinkpad E450

That's the only machine i have and so it getting heated up to the point i can'T work with it on my lap scared me i did not want to fry the board or CPU

That's when people stepped in giving suggestion and RT supported with his suggestions and question.

Now i am backing up all my stuff and reflashing, so i can report the findings and probably sort the issue.

re-insatlling* :D

installing*

Any other question you have please feel free to ask :)

Do you have thermal paste available to you?

nope

also it doesn't happen with linux

I don't even feel a bit of heat when i have linux installed

i don;t think its a hardware issue

or cooling paste wearing off

huh

?

you probably should replace the thermal paste anyway, but you're right - if it's not happening with another OS, it's a bug in a FreeBSD-specific component (namely loader)

Its a good suggestion  to replace the paste, but i am a bit worried about messing things up, as this is the only machine i've had since 2015 and will have going forward for a while. so i am being careful.

the xx50 lineup is quite old by now, and if you're on original paste that's a big risk. i replaced the paste in my t480 and it seemed to give me a free all-core overclock without tweaking any settings.

but that's outscope for here.

i will remember the paste stuff and may be get a professional handle it

it mught help a lot

I think there may be some kind of bug in the bootloader that's maybe making it infinite loop on the first core

goot to know about the paste stiff though

when you get into the OS does the system cool down?

that's what i am planin to figure with help

i amcurrently backing up eerythig so i can wipe and install freebsd to see whats the actual issue

no, i tried to check that mutiple times and the syste, stayed hot

which installer do you think would be best to get logs out of (may be through ssh, like in linux installation) dvd?cd?

TD can answer this too if he is here

now i am even scared to tag anyone :-p

:p

oh, if you need someone's attention specifically it's probably ok to tag them?

i know, but i can't figure out when is crossing the limit. I am usually very measured, even in real life, i hate bothering people or realying too much on someone.

but since you said he is busy with other stuff, i thoguht may be you know somethign i don't as i am new here and didn't want to bother him

but i believe i will be over this

:)

doesn't mean i will bother

TB said 13.2 would do, but mfsd image is only available for 13.1

mfsbsd*

TB?

RT sorry

yon he returns \o/ :P

wb

haha

conference?

buying food?

food indeed

Food is good.

better than no food, certainly

anyone here doing cross compilation for arm?

me

for armv7, anyway

nice.

(not actually right this second, but i have done it before and plan to do it again soon)

cross-compiling what? src or ports?

none of that. trying to build my own firmware.

oh.

I think we discussed this already

i got myself a teensy 4.1 which has a cortex m7. currently trying to figure this out. according to the teensy designer, i need the cores repository.

maybe, but if so last time was only hypothetical.

the cores thing fails compiling, apparently because of a syntax error in the includes from one of the arm-none-eabi-* packages. fails in stdlib.h in particular.

the error is?

1 sec, pasting

Title: stdin

__NULLABILITY_PRAGMA_PUSH seems to confuse it.

ok. so the main thing here I think is that something is assuming a hosted environment and that the host's include files are usable

hosted environment?

that symbol should be #defined to an empty value in cdefs.h

I'm not finding any cdefs.h

sys/cdefs.h

nope, sys only contains types.h which also doesn't defined the NULLABILITY_PRAGMA thing…

(these files are from arm-none-eabi-newlib btw.

the stdlib.h is clearly from freebsd, though?

phryk: oh, are you one of these people I caught an e-mail from complaining about newlib?

i'm really behind on taking a look at this

it worked the last time I seriously used it, but that was... right after I ported it, and I might've done backflips to make it work

kevans: i don't think so. not even completely sure what it is. i think it's an alternative libc?

yeah

RhodiumToad: no, look at the path in the paste, that's also from the arm-none-eabi-newlib package.

.../include-fixed

so it's the result of applying fixincludes to the system headers, most likely

so it still has freebsd-specific definitions in it, and without freebsd's sys/cdefs.h and so on, that won't work

you need a stdlib.h corresponding to what will be available on the target, not the host

RhodiumToad: huh? isn't that exactly what stdlib.h from the arm-none-eabi-newlib package is supposed to be?

haven't used that specific package

as i understand it, all the arm-none-eabi-* packages are for baremetal firmware without OS while arm-gnueabi-* stuff is for running linux on the µC.

i could be wrong on that, tho :P

does your .../include-fixed/stdlib.h have an #include <sys/cdefs.h> ?

it does, lol. so, that means the newlib package is just broken?

is there a sys/cdefs.h file anywhere in the package?

huh, there is.

/usr/local/arm-none-eabi/include/sys/cdefs.h

and does it differ from /usr/include/sys/cdefs.h ?

wildly.

hm.

is there a /usr/local/arm-none-eabi/include/stdlib.h ?

indeed there is. i guess now you want to know if/how that differs from the failing stdlib.h?

they do.

this is weird.

but yes, the -newlib package does legitimately include two stdlib.h files at different locations…

(checked with pkg query)

you likely need to tell it to ignore the /include-fixed/* ones.

maybe -nostdinc -isystem /usr/local/arm-none-eabi/include

i'm not even sure how they got in there. is there something like a make.conf or somesuch thing defines default locations for this specific compiler?

may also want -ffreestanding

well, i do want it to include the standard library… -isystem ignores a location? i assume that's a compiler flag?

-isystem PATH tells it to look for include files in PATH/* and treat them as system headers

you don't want it to include the host's standard library

that's not really what it's doing tho, is it? i'm not at all sure what this second stdlib.h is for, tho…^^

RT, If i need to check-for/take logs during installation, which image should i use for installation? DVD? I found that remote installation is possible through mfsbsd, but the image is only for 13.1. Your advise?

just install from dvd or memstick

okay, i would still be able to take logs and check stuffs to figure things out right?

I mean you can even install by hand via the memstick or dvd images by dropping into the shell

the reason i asked was that i used to install linux through ssh to get logs

I'm pretty sure that installation won't log anything useful for this issue that can't be determined more easily from the installed system

daemon oh is it? taht's cool. thanks :)

RT ah okay, fair enough

sozuba, youtube.com/watch?v=elWqsKXhNCE&t=7s I shot this a few years ago but it should still be about the same

Title: Howto: Install FreeBSD Manually - YouTube

daemon thats awesome :) Thanks will check it out

I really should do some more BSD videos :o

daemon would be great, peope like me would really benifit

benefit*

yeah its a pretty niche realm

I think other than me a couple of years ago the only real video poster is al

err I expected tab complte to do 'AllanJude' there

grmbl. trying to debug this and makefiles are just sooo garbage. deleted the only call to the compiler and replaced it with an (error …) and it still compiles stuff but doesn't show the error message. /o\

probably it is using implicit .c.o rules?

RhodiumToad: I don't think so. "all: $(TARGET).hex"

try this: make CC=echo

$(TARGET).hex depends on $(TARGET).elf, which in turn contains $(CC) <bunch of flags>

good idea

but does $(TARGET).elf depend on a bunch of .o files?

oh, right

$(OBJS) $(MCU_LD)

so if something depends on a .o file, and there's no explicit rule to build that file, it'll look for implicit rules to build it from other files, in this case the matching .c file

those implicit rules should use $(CC) as the compiler

indeed.

so, is there a way to first see what the current paths for includes and libraries are and then set them globally?

try /usr/local/bin/arm-none-eabi-gcc -print-search-dirs

er, no

adding -isystem just to CC fixes a lot of stuff already tho.

now it's failing because it includes Arduino.h (even tho I set NO_ARDUINO in my env), so I think this is now actually an issue with the upstream thing i'm trying to compile, not the toolchain. :)

okay, it's complaining about <type_traits>, which is a cpp thing. i think this might be a good point to pull the makefile out of there and start creating a minimal c-only build env.

so I think I have deleted my UEFI partition ages ago... this is the first computer I have owned that uses uefi. I still have a lenovo diagnostics thing I can boot into and a BIOS. Is this something I have to worry about?

are you booting from legacy bios?

what does gpart show tell you?

daemon: huh?