Re: [gentoo-user] Gentoo-sources : stable versions
I think its more complex than just falling behind - anything later gets the spectre fixes etc. and it appears not to be quite stable yet in some cases. I am on 4.9.95 for everything except a surface pro4 with 4.16.17 (as stable as anything can be on those things) which needs latest. BillK On 06/27/18 10:41, Philip Webb wrote: > 180626 Rich Freeman wrote: >> On Tue, Jun 26, 2018 at 8:58 PM Philip Webb wrote: >>> Does anyone know why the latest stable version of Gentoo-sources is 4.9.xx ? >>> I installed 4.9.16 , which I continue to use, on 2017-04-06 . >>> The tree contains versions of 4.14 4.16 4.17 , but all are still testing. >> I tend to just use my own upstream kernels. I'm following the 4.14 longterm >> and generally update within a few days of any release. >> That said, I have been burned by the odd regression. > Thanks for the other info (snipped). All Vanilla-sources are testing, > which seems to correspond to your "upstream" kernels. > > What does this say re recent kernel development or Gentoo's kernel team ? > -- very quick thanks as always to Gentoo's volunteer developers, > but something seems to be going astray here (smile). >
Re: [gentoo-user] Gentoo-sources : stable versions
180626 Rich Freeman wrote: > On Tue, Jun 26, 2018 at 8:58 PM Philip Webb wrote: >> Does anyone know why the latest stable version of Gentoo-sources is 4.9.xx ? >> I installed 4.9.16 , which I continue to use, on 2017-04-06 . >> The tree contains versions of 4.14 4.16 4.17 , but all are still testing. > I tend to just use my own upstream kernels. I'm following the 4.14 longterm > and generally update within a few days of any release. > That said, I have been burned by the odd regression. Thanks for the other info (snipped). All Vanilla-sources are testing, which seems to correspond to your "upstream" kernels. What does this say re recent kernel development or Gentoo's kernel team ? -- very quick thanks as always to Gentoo's volunteer developers, but something seems to be going astray here (smile). -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
Re: [gentoo-user] Gentoo-sources : stable versions
On Tue, Jun 26, 2018 at 8:58 PM Philip Webb wrote: > > Does anyone know why the latest stable version of Gentoo-sources is 4.9.xx ? > I installed 4.9.16 , which I continue to use, on 2017-04-06 . > The tree contains versions of 4.14 4.16 4.17 , but all are still testing. > I believe that some had been complaining about stability issues with 4.14, but personally I've been fine with it, and since I have a Ryzen CPU I want something that has the SMT support enabled. 4.9 is a longterm series, but the current upstream release is 4.9.110, and the Gentoo stable release is 4.9.95, so you seem to be pretty far behind. From a security standpoint a /recent/ 4.9 kernel should be fine. The gentoo patches do sometimes fix things and their curation might help you avoid some issues, but personally I tend to just use my own upstream kernels. I'm following the 4.14 longterm and generally update within a few days of any release. That said, I have been burned by the odd regression (a build last week actually was rebooting on me, which usually isn't something I have trouble with). Usually I migrate to a new longterm after it has been out for a few months, and only if it has zfs support. I don't know what the Gentoo plans are for 4.14 - I know they've been avoiding it for a while. I'm not sure if they ever plan to move to it, or if they're just hoping to skip it entirely. I think we should be getting a new longterm sometime in the next few months. -- Rich
[gentoo-user] Gentoo-sources : stable versions
Does anyone know why the latest stable version of Gentoo-sources is 4.9.xx ? I installed 4.9.16 , which I continue to use, on 2017-04-06 . The tree contains versions of 4.14 4.16 4.17 , but all are still testing. -- ,, SUPPORT ___//___, Philip Webb ELECTRIC /] [] [] [] [] []| Cities Centre, University of Toronto TRANSIT`-O--O---' purslowatchassdotutorontodotca
Re: [gentoo-user] Ethernet-over-USB confusion
On Tue, Jun 26, 2018 at 1:47 PM, wrote: > Hi, > > short question: > > I want to access my Raspberry Pi Zero W via Ethernet-over-USB > via ssh. > On the Raspberry I have to include a module called g_ether. > After rebooting I can see a new interface whith an IP assigned > via ifconfig. > On my PC I see a new inteface and dmesg tells me, that > usb0 was renamed to a hillarious cryoted other name > and this also shows up when using ifconfig. > But it has a complete different IP assigned, > > But I cannot ssh into the SoC even if the IP > address is corrected manually. > > What modules do I need loaded exactlu on the PC > and on the SoC? > > > Thanks for any help in advance! > Cheers > Meino > I haven't poked a pi zero yet with the usb ethernet gadget stuff, but I know the beaglebone side sets up its own dhcp server for the process, as per this in their getting started docs: "If connected via USB, a network adapter should show up on your computer. Your Beagle should be running a DHCP server that will provide your computer with an IP address of either 192.168.7.1 or 192.168.6.1, depending on the type of USB network adapter supported by your computer's operating system. Your Beagle will reserve 192.168.7.2 or 192.168.6.2 for itself." What IP/subnet are you seeing on the pi and your desktop/laptop it's connected to for the USB interfaces? -- Poison [BLX] Joshua M. Murphy
[gentoo-user] Re: Ethernet-over-USB confusion
On 2018-06-26 19:47, tu...@posteo.de wrote: > But I cannot ssh into the SoC even if the IP address is corrected > manually. Usually ssh is not the first thing you try when testing a new network setup. -- Please don't Cc: me privately on mailing lists and Usenet, if you also post the followup to the list or newsgroup. To reply privately _only_ on Usenet and on broken lists which rewrite From, fetch the TXT record for no-use.mooo.com.
Re: [gentoo-user] nodejs emerge fails
> On 2018-06-26, at 16:18, Jack wrote: > > On 2018.06.26 15:44, Christoph Böhmwalder wrote: >> On Tue, Jun 26, 2018 at 03:15:09PM -0400, Andrew Udvare wrote: >> > It should not be accessing this location in any case. What is the >> > environment like? What does env show? emerge --config output? >> > >> > It would almost seem like you have $HOME set to /home/christoph while >> > Portage is running (as root). >> The only thing that stands out about `env` is that XDG_CONFIG_HOME is set to >> /home/christoph/.config (running as root). I'm not sure if it's supposed to >> be this way, but I didn't find anything in my dotfiles that would suggest >> that I'm overwriting it. > That's a normal setting for you. The problem is that when you did sudo or su > to run emerge, it simply stayed in the environment, causing this problem. > When you su or sudo to run emerge (or just log in directly as root) you need > to be sure to do so in a way that cleans out your environment. I keep > thinking to file a bug to request emerge to sanitize the environment, or at > least add an option to make it do so. I often forget to use them, but I > created scripts (cleanemerge and cleanebuild) to do that for me. I had a similar issue with BC_ENV_ARGS being set while running Portage. If you have this set, some packages (I forget which) that use bc will try to read your .bcrc file and they will most likely fail. The scripts I used to use would unset this variable before calling emerge. I wrote Pezu ( https://github.com/Tatsh/pezu ) as a replacement for said scripts. It isolates the environment automatically since everything is run with the Subprocess module in Python (which does not inherit the environment by default). I run into less issues this way. Andrew
Re: [gentoo-user] nodejs emerge fails
> On 2018-06-26, at 15:44, Christoph Böhmwalder > wrote: > > On Tue, Jun 26, 2018 at 03:15:09PM -0400, Andrew Udvare wrote: >> It should not be accessing this location in any case. What is the >> environment like? What does env show? emerge --config output? >> >> It would almost seem like you have $HOME set to /home/christoph while >> Portage is running (as root). > > The only thing that stands out about `env` is that XDG_CONFIG_HOME is set to > /home/christoph/.config (running as root). I'm not sure if it's supposed > to be this way, but I didn't find anything in my dotfiles that would > suggest that I'm overwriting it. Use `su -` to become root so the environment you are coming from gets ignored. I actually have su aliased to `su -` because I rarely need the opposite functionality. Otherwise just `unset XDG_CONFIG_HOME` and other things that reference your user when you become root. You should only have these, if any at all: # env | fgrep XDG XDG_DATA_DIRS=/usr/local/share:/usr/share XDG_CONFIG_DIRS=/etc/xdg What it would seems to be is that npm uses configstore https://www.npmjs.com/package/configstore and since it sees XDG_CONFIG_HOME is set and is a real path it tries to write a file to there. It falls back to tmpdir otherwise which would be writable by Portage. https://github.com/yeoman/configstore/blob/master/index.js#L11 https://github.com/yeoman/configstore/blob/master/index.js#L32 > > $HOME points to /root. > > I'm not sure what you mean by "emerge --config output"? Running > `emerge --config` just gives: Sorry I meant `emerge --info`. Andrew
Re: [gentoo-user] nodejs emerge fails
On 2018.06.26 15:44, Christoph Böhmwalder wrote: On Tue, Jun 26, 2018 at 03:15:09PM -0400, Andrew Udvare wrote: > It should not be accessing this location in any case. What is the environment like? What does env show? emerge --config output? > > It would almost seem like you have $HOME set to /home/christoph while Portage is running (as root). The only thing that stands out about `env` is that XDG_CONFIG_HOME is set to /home/christoph/.config (running as root). I'm not sure if it's supposed to be this way, but I didn't find anything in my dotfiles that would suggest that I'm overwriting it. That's a normal setting for you. The problem is that when you did sudo or su to run emerge, it simply stayed in the environment, causing this problem. When you su or sudo to run emerge (or just log in directly as root) you need to be sure to do so in a way that cleans out your environment. I keep thinking to file a bug to request emerge to sanitize the environment, or at least add an option to make it do so. I often forget to use them, but I created scripts (cleanemerge and cleanebuild) to do that for me. $HOME points to /root. I'm not sure what you mean by "emerge --config output"? Running `emerge --config` just gives: # emerge --config nodejs Configuring pkg... Warning: ccache requested but no masquerade dircan be found in /usr/lib*/ccache/bin * pkg_config() is not defined: 'nodejs-8.11.1.ebuild' -- Regards, Christoph Jack
Re: [gentoo-user] Re: gcc failed to compile, mpfr-4 compiled successfully...
Am Montag, 25. Juni 2018, 17:17:31 CEST schrieb tu...@posteo.de: > > I am on glibc-(2.2) and the qfile command does not find that file. > Zong... you just voided your warranty. :) Seriously, please mention this *first* in the future when asking such questions. (And no, I dont know of a solution offhand. Might make sense to search gcc commit log...) -- Andreas K. Hüttel dilfri...@gentoo.org Gentoo Linux developer (council, toolchain, perl, libreoffice, comrel) signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] nodejs emerge fails
On Tue, Jun 26, 2018 at 03:15:09PM -0400, Andrew Udvare wrote: > It should not be accessing this location in any case. What is the environment > like? What does env show? emerge --config output? > > It would almost seem like you have $HOME set to /home/christoph while Portage > is running (as root). The only thing that stands out about `env` is that XDG_CONFIG_HOME is set to /home/christoph/.config (running as root). I'm not sure if it's supposed to be this way, but I didn't find anything in my dotfiles that would suggest that I'm overwriting it. $HOME points to /root. I'm not sure what you mean by "emerge --config output"? Running `emerge --config` just gives: # emerge --config nodejs Configuring pkg... Warning: ccache requested but no masquerade dircan be found in /usr/lib*/ccache/bin * pkg_config() is not defined: 'nodejs-8.11.1.ebuild' -- Regards, Christoph
Re: [gentoo-user] nodejs emerge fails
> On 2018-06-26, at 10:47, Christoph Böhmwalder > wrote: > > Hi, > > Lately I've been getting the following error while trying to emerge > nodejs: > > * Failed to set XATTR_PAX markings -me > /var/tmp/portage/net-libs/nodejs-8.11.1/image/usr/bin/node. This might be cause for concern. This does not look like it should fail. > * ACCESS DENIED: open_wr: > /home/christoph/.config/configstore/update-notifier-npm.json.331767419 > * ACCESS DENIED: unlink: > /home/christoph/.config/configstore/update-notifier-npm.JSON.331767419 It should not be accessing this location in any case. What is the environment like? What does env show? emerge --config output? It would almost seem like you have $HOME set to /home/christoph while Portage is running (as root). Andrew
Re: [gentoo-user] Ethernet-over-USB confusion
On Tue, Jun 26, 2018 at 12:47 PM, wrote:
> Hi,
>
> short question:
>
> I want to access my Raspberry Pi Zero W via Ethernet-over-USB
> via ssh.
> On the Raspberry I have to include a module called g_ether.
> After rebooting I can see a new interface whith an IP assigned
> via ifconfig.
> On my PC I see a new inteface and dmesg tells me, that
> usb0 was renamed to a hillarious cryoted other name
> and this also shows up when using ifconfig.
> But it has a complete different IP assigned,
>
> But I cannot ssh into the SoC even if the IP
> address is corrected manually.
>
Can you SSH into it at all via some other interface? Can you ping it?
Try g_multi so that you can have a serial terminal to log in with. If
possible switch to using g_ffs. I have included a script at the end of
my message which will create both CM and ACM (serial) endpoints.
On the device, bring up the interface and give a fixed static IP to
the interface. A route should be created but if not route everything
on that subnet through the gadget ethernet device.
On the host, bring up the interface and give a fixed static IP to the
interface. A route should be created but if not route everything on
that subnet through the gadget ethernet device.
It may be necessary to bring up the device first.
> What modules do I need loaded exactlu on the PC
> and on the SoC?
>
If usbX shows up and is renamed then you have the proper driver
loaded. There should be a message labelled "cdc_XXX." By default
g_ether should use CDC ECM (Ethernet control model), but it is also
possible to use CDC EEM (Ethernet emulation model), CDC NCM (network
control model), or RNDIS (remote network driver interface
specification, Windows).
Of the protocols ECM is the simplest, and in theory should have the
worst throughput do to inefficiency when wrapping Ethernet frames.
However in my tests it had the highest throughput by a few Mb/s
(~90Mbit). EEM and NCM were tied (88Mbit) though NCM is the newest and
may eventually surpass the others. RNDIS was slower by 10-15Mb/s
(70Mbit) and a full specification does not exist. However, it may be
necessary if you wish to easily use your device with Windows machines.
Cheers,
R0b0t1
---
#!/bin/bash
set -euo pipefail
# General configuration:
GADGET_NAME="testgadget"
LANGUAGE=0x409
MANUFACTURER="R0b0t1"
PRODUCT="Test Gadget"
# Function configuration:
HOST_ADDRESS="1a:55:89:a2:69:42"
DEV_ADDRESS="1a:55:89:a2:69:41"
if [ "$EUID" -ne 0 ]; then
echo "This script must run as root."
exit 1
fi
echo "Loading libcomposite..."
CONFIGFS="/sys/kernel/config/usb_gadget"
GADGET="$CONFIGFS/$GADGET_NAME"
modprobe libcomposite
while [ ! -d $CONFIGFS ]; do
sleep 0.01
done
echo "Cleaning existing gadget..."
find $GADGET/configs/*/* -maxdepth 0 -type l -exec rm {} \; || true
find $GADGET/configs/*/strings/* -maxdepth 0 -type d -exec rmdir {} \; || true
find $GADGET/os_desc/* -maxdepth 0 -type l -exec rm {} \; || true
find $GADGET/functions/* -maxdepth 0 -type d -exec rmdir {} \; || true
find $GADGET/strings/* -maxdepth 0 -type d -exec rmdir {} \; || true
find $GADGET/configs/* -maxdepth 0 -type d -exec rmdir {} \; || true
rmdir $GADGET || true
#find "$GADGET/configs" -maxdepth 2 -type l -exec "rm" {} \;
#find "$GADGET"/configs/*/strings/* -maxdepth 0 -type d -exec "rmdir" {} \;
#find "$GADGET"/os_desc/* -maxdepth 0 -type l -exec "rm" {} \;
#find "$GADGET"/configs/functions/* -maxdepth 0 -type d -exec "rmdir" {} \;
#find "$GADGET"/configs/* -maxdepth 0 -type d -exec "rmdir" {} \;
#rmdir "$GADGET"
echo "Creating gadget..."
mkdir "$GADGET"
cd "$GADGET"
echo "Configuring device identifiers..."
echo 0x1d6b > idVendor # Linux Foundation
echo 0x0104 > idProduct # Multifunction Composite Gadget
echo 0x0001 > bcdDevice # v0.1.0
echo 0x0200 > bcdUSB# USB 2.0
mkdir "strings/$LANGUAGE"
echo "$MANUFACTURER" > "strings/$LANGUAGE/manufacturer"
echo "$PRODUCT" > "strings/$LANGUAGE/product"
echo "Configuring composite device..."
echo 0xEF > bDeviceClass
echo 0x02 > bDeviceSubClass
echo 0x01 > bDeviceProtocol
# This is necessary for multifunction devices to work with Windows due
to deficiencies in the same's driver stack.
echo "Configuring OS-specific descriptors..."
echo 1 > os_desc/use
echo 0xcd> os_desc/b_vendor_code
echo MSFT100 > os_desc/qw_sign
echo "Creating Ethernet endpoint (CDC NCM) (1/1)..."
mkdir "functions/ncm.usb0"
#echo "$HOST_ADDRESS" > "functions/ncm.usb0/host_addr"
#echo "$DEV_ADDRESS" > "functions/ncm.usb0/dev_addr"
echo "Creating serial endpoint (1/2)..."
mkdir "functions/acm.usb0"
echo "Creating configuration..."
mkdir "configs/c.1"
echo 500 > "configs/c.1/MaxPower"
mkdir "configs/c.1/strings/$LANGUAGE"
echo "Config 1" > "configs/c.1/strings/$LANGUAGE/configuration"
ln -s "functions/ncm.usb0" "configs/c.1"
ln -s "functions/acm.usb0" "configs/c.1"
ln -s "configs/c.1" "os_desc/c.1"
echo "Attaching device..."
udevadm settle
ls "/sys/class/udc/" > UDC
echo "Done."
[gentoo-user] Ethernet-over-USB confusion
Hi, short question: I want to access my Raspberry Pi Zero W via Ethernet-over-USB via ssh. On the Raspberry I have to include a module called g_ether. After rebooting I can see a new interface whith an IP assigned via ifconfig. On my PC I see a new inteface and dmesg tells me, that usb0 was renamed to a hillarious cryoted other name and this also shows up when using ifconfig. But it has a complete different IP assigned, But I cannot ssh into the SoC even if the IP address is corrected manually. What modules do I need loaded exactlu on the PC and on the SoC? Thanks for any help in advance! Cheers Meino
[gentoo-user] nodejs emerge fails
Hi, Lately I've been getting the following error while trying to emerge nodejs: >>> Install nodejs-8.11.1 into /var/tmp/portage/net-libs/nodejs-8.11.1/image/ >>> category net-libs make -j9 -l8 install DESTDIR=/var/tmp/portage/net-libs/nodejs-8.11.1/image/ make -C out BUILDTYPE=Release V=1 touch 95791f913e60cc9fcb29d34b6a73660b74906c54.intermediate LD_LIBRARY_PATH=/var/tmp/portage/net-libs/nodejs-8.11.1/work/node-v8.11.1/out/Release/lib.host:/var/tmp/portage/net-libs/nodejs-8.11.1/work/node-v8.11.1/out/Release/lib.target:$LD_LIBRARY_PATH; export LD_LIBRARY_PATH; cd ../deps/v8/src/inspector; mkdir -p /var/tmp/portage/net-libs/nodejs-8.11.1/work/node-v8.11.1/out/Release/obj/gen/src/inspector/protocol /var/tmp/portage/net-libs/nodejs-8.11.1/work/node-v8.11.1/out/Release/obj/gen/include/inspector; python ../../third_party/inspector_protocol/CodeGenerator.py --jinja_dir ../../third_party --output_base "/var/tmp/portage/net-libs/nodejs-8.11.1/work/node-v8.11.1/out/Release/obj/gen/src/inspector" --config inspector_protocol_config.json rm 95791f913e60cc9fcb29d34b6a73660b74906c54.intermediate if [ ! -r node -o ! -L node ]; then ln -fs out/Release/node node; fi /usr/bin/python2.7 tools/install.py install '/var/tmp/portage/net-libs/nodejs-8.11.1/image/' '/usr' * PT_PAX marking -m /var/tmp/portage/net-libs/nodejs-8.11.1/image/usr/bin/node with paxctl * PT_PAX marking -m /var/tmp/portage/net-libs/nodejs-8.11.1/image/usr/bin/node with scanelf * XATTR_PAX marking -me /var/tmp/portage/net-libs/nodejs-8.11.1/image/usr/bin/node with setfattr * Failed to set XATTR_PAX markings -me /var/tmp/portage/net-libs/nodejs-8.11.1/image/usr/bin/node. * ACCESS DENIED: open_wr: /home/christoph/.config/configstore/update-notifier-npm.json.331767419 * ACCESS DENIED: unlink: /home/christoph/.config/configstore/update-notifier-npm.json.331767419 ┌─┐ │ npm update check failed │ │ Try running with sudo or get access │ │to the local update config store via │ │ sudo chown -R $USER:$(id -gn $USER) /home/christoph/.config │ └─┘ >>> Completed installing nodejs-8.11.1 into >>> /var/tmp/portage/net-libs/nodejs-8.11.1/image/ * Final size of build directory: 302712 KiB (295.6 MiB) * Final size of installed tree: 44524 KiB ( 43.4 MiB) * --- ACCESS VIOLATION SUMMARY --- * LOG FILE: "/var/log/sandbox/sandbox-26618.log" * VERSION 1.0 FORMAT: F - Function called FORMAT: S - Access Status FORMAT: P - Path as passed to function FORMAT: A - Absolute Path (not canonical) FORMAT: R - Canonical Path FORMAT: C - Command Line F: open_wr S: deny P: /home/christoph/.config/configstore/update-notifier-npm.json.331767419 A: /home/christoph/.config/configstore/update-notifier-npm.json.331767419 R: /home/christoph/.config/configstore/update-notifier-npm.json.331767419 C: npm F: unlink S: deny P: /home/christoph/.config/configstore/update-notifier-npm.json.331767419 A: /home/christoph/.config/configstore/update-notifier-npm.json.331767419 R: /home/christoph/.config/configstore/update-notifier-npm.json.331767419 C: npm * I've set the permissions on the ~/.config/configstore directory to 777 for testing purposes, and it still doesn't work. Anyone have any ideas? -- Regards, Christoph
[gentoo-user] Re: gcc failed to compile, mpfr-4 compiled successfully...
On 25/06/18 18:17, tu...@posteo.de wrote: On 06/25 12:33, Nikos Chantziaras wrote: Is the file there? It's "/usr/include/sys/ustat.h" and qfile says it's from glibc: $ qfile /usr/include/sys/ustat.h sys-libs/glibc (/usr/include/sys/ustat.h) I'm on glibc-2.27-r4. If you can't get rid of the error, it might be worth trying disabling the "sanitize" USE of gcc. I am on glibc- OK, I thought you were being serious. Never mind.
