cnorthway has quit [Remote host closed the connection]
cnorthway has joined #openwrt-devel
cnorthway has quit [Ping timeout: 480 seconds]
Tapper has quit [Ping timeout: 480 seconds]
lynxis has quit [Remote host closed the connection]
goliath has quit [Quit: SIGSEGV]
Tapper has joined #openwrt-devel
<djfe>
bye for now
<djfe>
will try again tomorrow morning now worries, I know I'm up way too late lol
<djfe>
s/now/no/
djfe has quit [Quit: Page closed]
Mangix has quit [Read error: No route to host]
Mangix has joined #openwrt-devel
<yolo>
for ucode the name uc is taken by unreal engine's script in vim, it actually highlights well for openwrt uc file, is there a better syntax highlight for it?
tSYS has quit [Quit: *squeak*]
tSYS has joined #openwrt-devel
minimal has quit [Quit: Leaving]
<dwfreed>
yolo: I doubt anybody's bothered to write a vim syntax hilighting script for it yet
<dwfreed>
but it is js inspired, so that might work
cmonroe has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
cmonroe has joined #openwrt-devel
cmonroe has quit []
paper_ is now known as paper__
paper_ has joined #openwrt-devel
<Znevna>
:(
<Znevna>
there's no ubi resize option?:P
<robimarko>
Well, how can you resize it on the fly if the partition table changed?
<Znevna>
it somehow boots if you keep the settings when doing the sysupgrade
<Znevna>
and I can see the new size in the bootlog
<Znevna>
but .. we'd have to resize the actual volume(?) somehow
<yolo>
will ucode also 'replace' config_foreach/config_get/config_set/option_cb, it seems trying to replace lua and shell all the way
djfe has joined #openwrt-devel
<djfe>
I wanna help clean up stale/resolved issues on Github, I already got like 2 or 3 of them. Whom can I contact to get them closed? Or is it possible to flag such issues with the bot?
<Znevna>
eh? what issues? :P
<Znevna>
hello
<hauke>
djfe: Please add a comment into the issue
<hauke>
with some details like wheer it was fixed
djfe has quit [Remote host closed the connection]
<Mangix>
hrm coverity email. interesting
<Mangix>
hauke: are patchwork patches still getting merged?
<slh>
Znevna: do I read it correctly that we're talking about a 128 MB flash, with 2*(4+46) MB dual-firmware? if so, *I* wouldn't bother, at all. 46 MB is plenty for OpenWrt and dual-firmware halves your flash anyways (I'd rather spend efforts to actually get the dual-firmware feature working, than messing with the partitioning). <20 MB usable space would be another question, but 4+46 MB really is plenty
<Znevna>
it's not actually dual firmware capable
<robimarko>
It would break compat with current versions
<Znevna>
the 2nd firmware is only used by uboot in case the primary firmware is corrupted
<robimarko>
Though, in theory since kernel is separate it might actually work
<robimarko>
But it requires ubiformating that new partition
<robimarko>
So you cant just sysupgrade
<slh>
that's odd, but still, with 4+46 MB, the the risks would imho outweigh the benefit
<Znevna>
and by using it I mean it copies the 2nd firmware to the first partition and boots the first partition
<Znevna>
robimarko losing data is no biggie
<Znevna>
if we can avoid a bootlooper
<slh>
especially as Asus (based on my experiences with the map-ac2200) does not have a reliable push-button tftp recovery (it's there, but doesn't work with 'unexpected' ubi contents, so unusable for desaster recovery, you have to explicitly invalidate the ubi)
<Znevna>
going back to stock with asus recovery is easy on this one
<slh>
(in other words, serial console required for recovering)
<Znevna>
can be triggered by button
<robimarko>
Znevna: You cant avoid bootloping
<slh>
greta itf that works
<robimarko>
That is why in previous cases of breaking the partition table you have to increase the COMPAT_VERSION to avoid easily sysupgrading
<robimarko>
As it will soft brick it
<Znevna>
robimarko: I swear I got it to boot at least twice with a sysupgrade (keeping data) obviously with the same ubisize (tho bootlog showed the larger size)
<Znevna>
but it would go into bootloop on a reset
<Znevna>
I wonder what this sysupgrade flag is supposed to do: -p do not attempt to restore the partition table after flash.
<robimarko>
I am honestly surprised that UBI mounted properly
<robimarko>
As like more than half of your partition will be non UBI data
<robimarko>
BTW, has anybody ever implemented PSCI via U-Boot serving as EL3?
<robimarko>
I can get spin-tables working great, but if trying to do the same via PSCI it fails to bringup the core
<Znevna>
I can't reproduce that behaviour tho, probably because of some ubi remnants on nand
<robimarko>
Honestly, usually breaking compat version isnt something that is done unless we are talking about partitions being too small
<robimarko>
Like, unusable small
Ryncewynd has quit [Read error: Connection reset by peer]
robimarko has quit [Quit: Leaving]
philipp64 has joined #openwrt-devel
<slh>
Znevna: just to add some perspective, http://paste.debian.net/hidden/56bc8451/ (140 MB fs overhead for f2fs is insane, as you see the actual file size on the overlay is under 0.2 MB; and those 18 MB on /srv/ are just static firmware files for my Cisco SIP desk phones); 32 GB SSD
<Znevna>
I'm not a fan of doing this if it wasn't done before the device PR got accepted either
<Znevna>
just that there were a few PRs trying to do the same and got stuck
<Znevna>
I'm just trying to either find a solution or just to add a good enough reason to the current PR that we shouldn't do this.
<slh>
NACK, high risk for little gain ;)
<dwfreed>
f2fs overhead is really ridiculous
<dwfreed>
like, who thought that was a good idea
hanetzer has joined #openwrt-devel
cmonroe has quit [Ping timeout: 480 seconds]
cmonroe has joined #openwrt-devel
<slh>
yeah, it's insane - and especially on the x86 images totally unusable. 104 MB, so just over the threshold for f2fs to be default, but that means only a few MB remain usable after the f2fs fs overhead
<Borromini>
i suppose that gets better once you enter gigabyte territory?
<slh>
Borromini: the problem is the default image for x86 being set 104 MB. I do build for myself and set the kernel to 32 MB and the rootfs to 960 MB, still, I have almost 150 MB useless fs overhead with f2fs (as I have enough unused space on my SSD, *I* don't really care, but f2fs is totally braindead)
<slh>
Borromini: /dev/loop0 899.9M 144.2M 755.7M 16% /overlay <-- with 165 KB used in files on the overlay
<dwfreed>
Borromini: not really
<dwfreed>
I mean, yes, as a total percentage of available space, it does get better, but the overhead of f2fs is still insane compared to *any* other filesystem
<stintel>
maybe they want to be absolutely sure not to follow in btrfs' footsteps with unrecoverable -ENOSPC ¯\_(ツ)_/¯
<dwfreed>
rofl
<slh>
the fs overhead for ext4 on my data partition (remaining space on that 32 GB SSD, 28.3 GB) is under 100 KB