View previous topic :: View next topic |
Author |
Message |
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Sun Apr 14, 2019 8:05 pm Post subject: [solved] dhcpcd: can't get an ip from my router |
|
|
hi men,
i have troubles to make my wired network working. i've spent a couple of weeks trying many configs changes but without success. my eth iface always gets 169.254.44.99 (IPv4LL, see a log below). when i try to set a static ip addr i'm even not able to ping my router. my firewall is not a problem as i've tried to switch it off but the behavior of dhcpcd is the same. i've tried a different eth card (reliable INTEL PRO/1000 CT pcie) but hasn't helped... :-/
router works great with my old pc with gentoo installed (same dhcp and net config). also when i boot systemrescuecd (last release with gentoo) it works great as well so router, cable, network and pc are ok i'm pretty sure. time to time during reboot it gets ip address and works but it happens very seldom. also time to time dhcpcd automaticaly reloads a request (without any changes) itself and gets the right ip addr but it takes about 2 hours (it happend twice). i think there might be a problem with my kernel configuration but i'll really appreciate your help because maybe i'm wrong...
my hw: router TP-LINK Archer C5400, MB Asus PRIME X299-A, eth I219-V (on MB), CPU i9-7940X
dmesg | grep e1000e
Code: |
[ 6.460474] e1000e: Intel(R) PRO/1000 Network Driver - 3.2.6-k
[ 6.460475] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[ 6.462457] e1000e 0000:00:1f.6: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[ 6.712848] e1000e 0000:00:1f.6 0000:00:1f.6 (uninitialized): registered PHC clock
[ 6.793519] e1000e 0000:00:1f.6 eth0: (PCI Express:2.5GT/s:Width x1) 4c:ed:fb:42:74:c5
[ 6.793520] e1000e 0000:00:1f.6 eth0: Intel(R) PRO/1000 Network Connection
[ 6.793571] e1000e 0000:00:1f.6 eth0: MAC: 12, PHY: 12, PBA No: FFFFFF-0FF
[ 6.795934] e1000e 0000:00:1f.6 enp0s31f6: renamed from eth0
|
when i try to get an ip addr manually using dhcpcd after reboot it always finishes just like that:
dhcpcd -kn -t 10 -w --waitip=4 -d enp0s31f6
Code: |
sending signal HUP to pid 6683
dhcpcd-7.1.1 starting
enp0s31f6: executing `/lib/dhcpcd/dhcpcd-run-hooks' PREINIT
enp0s31f6: executing `/lib/dhcpcd/dhcpcd-run-hooks' CARRIER
DUID 00:01:00:01:1a:7d:2a:07:00:1c:c0:4c:41:c9
enp0s31f6: IAID fb:42:74:c5
enp0s31f6: delaying IPv4 for 0.3 seconds
enp0s31f6: soliciting a DHCP lease
enp0s31f6: sending DISCOVER (xid 0x8e16232c), next in 4.8 seconds
enp0s31f6: sending DISCOVER (xid 0x8e16232c), next in 7.7 seconds
enp0s31f6: using IPv4LL address 169.254.46.99
enp0s31f6: adding route to 169.254.0.0/16
enp0s31f6: adding default route
enp0s31f6: ARP announcing 169.254.46.99 (1 of 2), next in 2.0 seconds
enp0s31f6: executing `/lib/dhcpcd/dhcpcd-run-hooks' IPV4LL
forking to background
forked to background, child pid 7414
|
my configs: openrc (no systemd), eudev, no selinux, kernel compiled using genkernel
kernel gentoo-sources-4.19.27-r1
lsmod
dmesg
cat /etc/conf.d/net
Code: |
#
# lo
#
dns_domain_lo="homenetwork"
#
# enp0s31f6 (int. intel)
#
config_enp0s31f6="dhcp"
dhcpd_enp0s31f6="-t 5"
dhcp_enp0s31f6="nonis"
#enable_ipv6_enp0s31f6="false"
dns_domain_enp0s31f6="homenetwork"
#
fallback_enp0s31f6="192.168.1.3/24"
fallback_route_enp0s31f6="default via 192.168.1.1"
|
cat /etc/dhcpcd.conf | grep -v ^# | grep -v ^$
Code: |
hostname
duid
persistent
option rapid_commit
option domain_name_servers, domain_name, domain_search, host_name
option classless_static_routes
option interface_mtu
require dhcp_server_identifier
slaac private
timeout 60
|
i'd say that eth card is not supported by e1000e but systemrescuecd uses the same module and it works like a charm :-/
does anybody has experience with the same eth card on that MB?
Last edited by JohnBlbec on Thu Apr 18, 2019 11:22 am; edited 2 times in total |
|
Back to top |
|
|
UberLord Retired Dev
Joined: 18 Sep 2003 Posts: 6835 Location: Blighty
|
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Mon Apr 15, 2019 12:51 pm Post subject: |
|
|
UberLord wrote: | Try the usual suspects - comment out rapid_commit, swap clientid for duid.
But basically your router is not replying to your DHCP request. |
well, it hasn't helped. no change.
i've noticed ethtool says "Link detected: no" and i don't know why :-(
it looks like it is not able to detect/set speed and duplex mode because it says unknow or error. i've tried to set it manually using ethtool -s enp0s31f6 speed 1000 duplex full autoneg on but it changes nothing...
wtf why sytemrescuecd works ok with the same hw and cable?
Last edited by JohnBlbec on Mon Apr 15, 2019 1:19 pm; edited 1 time in total |
|
Back to top |
|
|
UberLord Retired Dev
Joined: 18 Sep 2003 Posts: 6835 Location: Blighty
|
|
Back to top |
|
|
Jaglover Watchman
Joined: 29 May 2005 Posts: 8291 Location: Saint Amant, Acadiana
|
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Mon Apr 15, 2019 1:20 pm Post subject: |
|
|
Jaglover wrote: | Quote: | i've noticed ethtool says "Link detected: no" and i don't know why :-( |
This is your problem, indeed. Do you have more than one ethernet port? |
as i wrote above i've tested also intel 1000/pro eth card inserted in pcie but w/o success. that's why i think the problem might be in the kernel settings... |
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Mon Apr 15, 2019 1:22 pm Post subject: |
|
|
UberLord wrote: | Code: | dhcp_enp0s31f6="-t 5"
dhcp_enp0s31f6="nonis" |
That part of your config doesn't look correct. |
this part is correct as i'm using it on that pc which i'm using for writing this message and it works years like a charm!
what do you think is wrong?
[edited]
yes there is a typo but only here in the forum: dhcp_enp0s31f6="-t 5" should be dhcpd_enp0s31f6="-t 5" (fixed)
Last edited by JohnBlbec on Mon Apr 15, 2019 1:31 pm; edited 4 times in total |
|
Back to top |
|
|
Jaglover Watchman
Joined: 29 May 2005 Posts: 8291 Location: Saint Amant, Acadiana
|
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Mon Apr 15, 2019 1:25 pm Post subject: |
|
|
Jaglover wrote: | So you say it is possible the cable is connected to the wrong port? :? |
no it is not possible because the MB has the only eth port and the same port and cable works with systemrescuecd and also in w10.
and btw. the eth socket has two led lights - green and orange - both are on. green means connection is ok and orange means 1000 speed even ethtool says link not detected.
(hey men, i've been using gentoo since 2003 so i'm really not a beginner who is able to insert cable to a different port)
Last edited by JohnBlbec on Mon Apr 15, 2019 1:55 pm; edited 1 time in total |
|
Back to top |
|
|
UberLord Retired Dev
Joined: 18 Sep 2003 Posts: 6835 Location: Blighty
|
Posted: Mon Apr 15, 2019 1:44 pm Post subject: |
|
|
JohnBlbec wrote: | UberLord wrote: | Code: | dhcp_enp0s31f6="-t 5"
dhcp_enp0s31f6="nonis" |
That part of your config doesn't look correct. |
this part is correct as i'm using it on that pc which i'm using for writing this message and it works years like a charm!
what do you think is wrong?
[edited]
yes there is a typo but only here in the forum: dhcp_enp0s31f6="-t 5" should be dhcpd_enp0s31f6="-t 5" (fixed) |
dhcpcd_enp0s31f6="-t 5" _________________ Use dhcpcd for all your automated network configuration needs
Use dhcpcd-ui (GTK+/Qt) as your System Tray Network tool |
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Mon Apr 15, 2019 2:00 pm Post subject: |
|
|
UberLord wrote: | JohnBlbec wrote: | UberLord wrote: | Code: | dhcp_enp0s31f6="-t 5"
dhcp_enp0s31f6="nonis" |
That part of your config doesn't look correct. |
this part is correct as i'm using it on that pc which i'm using for writing this message and it works years like a charm!
what do you think is wrong?
[edited]
yes there is a typo but only here in the forum: dhcp_enp0s31f6="-t 5" should be dhcpd_enp0s31f6="-t 5" (fixed) |
dhcpcd_enp0s31f6="-t 5" |
you're right! just now i'm using "dhcpd_..." but you can believe me i've tested also "dhcpcd_...". have a look at my config for the mentioned pcie intel 1000/pro card which i've bought because i thought my problem is the internal eth card I219-V :-(
Code: |
#
# enp179s0 (ext. intel)
#
config_enp179s0="dhcpcd"
modules_enp179s0="dhcpcd"
dhcp_enp179s0="release nonis"
dhcpcd_enp179s0="--timeout 10 --ipv4only --noipv4ll"
enable_ipv6_enp179s0="false"
dns_domain_enp179s0="homenetwork"
#
fallback_enp179s0="192.168.1.4/24"
fallback_route_enp179s0="default via 192.168.1.1"
|
but what seems strange to me is one row from dmesg. there is something "uninitialized" and there is also missing very important info that the link is up!
this is the internal problematic eth in my new pc:
[ 6.460474] e1000e: Intel(R) PRO/1000 Network Driver - 3.2.6-k
[ 6.460475] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[ 6.462457] e1000e 0000:00:1f.6: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[ 6.712848] e1000e 0000:00:1f.6 0000:00:1f.6 (uninitialized): registered PHC clock
[ 6.793519] e1000e 0000:00:1f.6 eth0: (PCI Express:2.5GT/s:Width x1) 4c:ed:fb:42:74:c5
[ 6.793520] e1000e 0000:00:1f.6 eth0: Intel(R) PRO/1000 Network Connection
[ 6.793571] e1000e 0000:00:1f.6 eth0: MAC: 12, PHY: 12, PBA No: FFFFFF-0FF
[ 6.795934] e1000e 0000:00:1f.6 enp0s31f6: renamed from eth0
and this is the same dmesg ouput from my old working pc (notice last line!)
[ 12.594133] e1000e: Intel(R) PRO/1000 Network Driver - 3.2.6-k
[ 12.594133] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[ 12.594353] e1000e 0000:00:19.0: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[ 12.891169] e1000e 0000:00:19.0 eth0: (PCI Express:2.5GT/s:Width x1) 00:1c:c0:4c:41:c9
[ 12.891171] e1000e 0000:00:19.0 eth0: Intel(R) PRO/1000 Network Connection
[ 12.891190] e1000e 0000:00:19.0 eth0: MAC: 7, PHY: 6, PBA No: FFFFFF-0FF
[ 13.576714] e1000e 0000:00:19.0 enp0s25: renamed from eth0
[ 115.985853] e1000e: enp0s25 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx
so from some reason my eth card in my new pc is not able to come up :-( |
|
Back to top |
|
|
UberLord Retired Dev
Joined: 18 Sep 2003 Posts: 6835 Location: Blighty
|
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Mon Apr 15, 2019 6:16 pm Post subject: |
|
|
sound like a good idea... i've tried it. i've created a file "/etc/modprobe.d/e1000e.conf" and added a line "options e1000e SmartPowerDownEnable=1" because 0 is a default. unfortunately, it hasn't helped and link is still not up :-( i'm also not able to stop or restart network using "/etc/init.d/net.enp0s31f6 stop" - it always stuck :-( i've tried to modprobe -r e1000e but it stuck too. there must be wrong something in the kernel configuration :-(
Last edited by JohnBlbec on Mon Apr 15, 2019 6:49 pm; edited 1 time in total |
|
Back to top |
|
|
1clue Advocate
Joined: 05 Feb 2006 Posts: 2569
|
Posted: Mon Apr 15, 2019 6:49 pm Post subject: |
|
|
Can you check sysrescuecd boot, check if it's configured at full speed or slower, and if it's slower then whether it's full duplex.
Also check the name of the interface and the mac address on both.
I know you've been at this for awhile, so have I and I'm sure some of these other guys too. Sometimes you need a reminder to find out if it's plugged in.
So just a reminder that an ethernet cable can have 1 wire broken and still function at a lower speed, or seem to function 1 direction but not the other.
Also when I was messing with interface names I messed up one of the files, and had to move it to another directory to disable the entire file, to figure out what file the configuration problem was in. I basically had to go back to a default network setup in order to get my system working again. Once I figured out what file was broken I could narrow the search. |
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Mon Apr 15, 2019 7:17 pm Post subject: |
|
|
first of all i really appreciate your help, guys. thanks for your patience, advice and answers...
1clue wrote: | Can you check sysrescuecd boot, check if it's configured at full speed or slower, and if it's slower then whether it's full duplex. |
yes i'll paste ethtool output tomorrow but i've checked it in w10 and the speed was 1000. the mac address'll be the same i guess because i always need from dhcp the same ip address based on my mac address...
1clue wrote: | Also check the name of the interface and the mac address on both. |
yes i'll check it too and let you know.
1clue wrote: | I know you've been at this for awhile, so have I and I'm sure some of these other guys too. Sometimes you need a reminder to find out if it's plugged in. |
i see and you're right. i didn't want to offend anybody. i just want to give you a notice and nothing more. i can overlooked something, of course...
1clue wrote: | So just a reminder that an ethernet cable can have 1 wire broken and still function at a lower speed, or seem to function 1 direction but not the other. |
you're right and i'll test it again but i've tried 3 various cables to be sure because i spent egregious much money for a new pc and i can't use it for a long time because network troubles so i'm quite nerves, you know...
1clue wrote: | Also when I was messing with interface names I messed up one of the files, and had to move it to another directory to disable the entire file, to figure out what file the configuration problem was in. I basically had to go back to a default network setup in order to get my system working again. Once I figured out what file was broken I could narrow the search. |
i see. i've recompiled and configured my kernel maybe 100 times and played with various configs as well before i decided to write here for a help... |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54237 Location: 56N 3W
|
Posted: Mon Apr 15, 2019 8:03 pm Post subject: |
|
|
JohnBlbec,
A you read, keep in minhd that there are no silly questions, except tha one that is never asked.
JohnBlbec wrote: | i see. i've recompiled and configured my kernel maybe 100 times and played with various configs as well before i decided to write here for a help... |
Are you runinig the kernel you think you are?
What does uname -a say?
Code: | $ uname -a
Linux NeddySeagoon_Static 5.0.2-gentoo #2 SMP PREEMPT Sun Mar 17 18:30:20 GMT 2019 x86_64 AMD Phenom(tm) II X6 1090T Processor AuthenticAMD GNU/Linux |
Sun Mar 17 18:30:20 GMT 2019 is the build time of the running kernel, does the build time of your runnig kernel look correct?
The content of /etc/modprobe.d/ is only applied to loadable modules. If your e1000e is built into the kernel, it will have no effect, since the code will be initalised before root is mounted.
Post your output and put your kernel .config file onto a pastebin site, post the link so we can find it. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
1clue Advocate
Joined: 05 Feb 2006 Posts: 2569
|
Posted: Mon Apr 15, 2019 10:47 pm Post subject: |
|
|
JohnBlbec wrote: | ...
1clue wrote: | I know you've been at this for awhile, so have I and I'm sure some of these other guys too. Sometimes you need a reminder to find out if it's plugged in. |
i see and you're right. i didn't want to offend anybody. i just want to give you a notice and nothing more. i can overlooked something, of course...
|
And we're not trying to offend you either. I at least can tell you're not a novice from the way you're talking, but when we ask the "is it plugged in" questions we're trying to eliminate the obvious, because so far everything I can think of has been covered here.
Quote: |
1clue wrote: | So just a reminder that an ethernet cable can have 1 wire broken and still function at a lower speed, or seem to function 1 direction but not the other. |
you're right and i'll test it again but i've tried 3 various cables to be sure because i spent egregious much money for a new pc and i can't use it for a long time because network troubles so i'm quite nerves, you know...
|
OK so you've eliminated cables from the equation. Have you tried a different port on your switch? Sometimes the switch ports can get flaky. This seems unlikely to be the problem since the same port works at full speed in Windows or sysrescuecd. Just something to check off the list.
Quote: |
1clue wrote: | Also when I was messing with interface names I messed up one of the files, and had to move it to another directory to disable the entire file, to figure out what file the configuration problem was in. I basically had to go back to a default network setup in order to get my system working again. Once I figured out what file was broken I could narrow the search. |
i see. i've recompiled and configured my kernel maybe 100 times and played with various configs as well before i decided to write here for a help... |
This might not be a kernel thing. I horsed around with it for awhile but found out mine was an error in /etc/udev/rules.d/70-net-name-use-custom.rules. My system has 7 NICs and I had a bunch of changes for each of them here, trying to make a KVM host with a router controlling the physical NICs. If the file is not present that's no big deal, I created it in order to get non-standard behavior. If you have a file there doing anything with networking you may want to move it out of there for awhile to reduce the number of places where there may be an error.
Have you compared the kernel config from the sysrescuecd kernel and yours? I know that seems like a huge PITA but maybe you can just do networking features. |
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Tue Apr 16, 2019 6:54 pm Post subject: |
|
|
hi guys,
i've just finished my regular work finally so i can go in to get information you wanted yesterday...
booted systemrescuecd using the same cable, eth and router port
Code: |
# uname -a
Linux sysresccd 4.14.70-std531-amd64 #2 SMP Sat Sep 15 16:16:19 UTC 2018 x86_64 Intel(R) Core(TM) i9-7940X CPU @ 3.10GHz GenuineIntel GNU/Linux
|
here you can find systemrescudecd kernel config
here is lspci -knn output
you can find my own kernel config in my the 1st message in this thread but i can paste it again here kernel gentoo-sources-4.19.27-r1. i haven't tried to compare it to the systemrescuecd config because it is not so easy but i've tried to compile a new kernel based on the systemrescuecd kernel. it took much time to change everything what i need to change and the result was not very stable so i return back to my own. btw. a kernel config which i'm using on my old pc (with working network) is also quite optimized and you can find it here old pc kernel-config-x86_64-4.14.105-gentoo-r1. i've also tried to build a new one based on it but the new hardware is soo different... i gave it up and rather started from scratch (e1000e is compiled as a module and there is only 1 kernel so it is not possible to boot another one).
Code: |
ifconfig enp0s31f6
enp0s31f6: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
inet 192.168.1.3 netmask 255.255.255.0 broadcast 192.168.1.255
inet6 fe80::5565:eb83:f013:2755 prefixlen 64 scopeid 0x20<link>
ether 4c:ed:fb:42:74:c5 txqueuelen 1000 (Ethernet)
RX packets 6400 bytes 7424072 (7.0 MiB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 4620 bytes 467197 (456.2 KiB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
device interrupt 20 memory 0x92f00000-92f20000
|
Code: |
# ethtool enp0s31f6
Settings for enp0s31f6:
Supported ports: [ TP ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Supported pause frame use: No
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
1000baseT/Full
Advertised pause frame use: No
Advertised auto-negotiation: Yes
Speed: 1000Mb/s
Duplex: Full
Port: Twisted Pair
PHYAD: 1
Transceiver: internal
Auto-negotiation: on
MDI-X: on (auto)
Supports Wake-on: pumbg
Wake-on: g
Current message level: 0x00000007 (7)
drv probe link
Link detected: yes
|
Code: |
# dmesg | grep e1000e
[ 155.051470] e1000e: Intel(R) PRO/1000 Network Driver - 3.2.6-k
[ 155.051471] e1000e: Copyright(c) 1999 - 2015 Intel Corporation.
[ 155.053743] e1000e 0000:00:1f.6: Interrupt Throttling Rate (ints/sec) set to dynamic conservative mode
[ 155.253037] e1000e 0000:00:1f.6 0000:00:1f.6 (uninitialized): registered PHC clock
[ 155.388700] e1000e 0000:00:1f.6 eth0: (PCI Express:2.5GT/s:Width x1) 4c:ed:fb:42:74:c5
[ 155.388703] e1000e 0000:00:1f.6 eth0: Intel(R) PRO/1000 Network Connection
[ 155.388754] e1000e 0000:00:1f.6 eth0: MAC: 12, PHY: 12, PBA No: FFFFFF-0FF
[ 155.398576] e1000e 0000:00:1f.6 enp0s31f6: renamed from eth0
[ 166.910856] e1000e: enp0s31f6 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx
|
Code: |
# lsmod
Module Size Used by
binfmt_misc 16384 1
ses 16384 0
enclosure 16384 1 ses
intel_rapl 24576 0
nouveau 1421312 2
x86_pkg_temp_thermal 16384 0
intel_powerclamp 16384 0
eeepc_wmi 16384 0
coretemp 16384 0
asus_wmi 28672 1 eeepc_wmi
i2c_algo_bit 16384 1 nouveau
crct10dif_pclmul 16384 0
ttm 77824 1 nouveau
sparse_keymap 16384 1 asus_wmi
drm_kms_helper 131072 1 nouveau
rfkill 24576 1 asus_wmi
drm 311296 5 drm_kms_helper,ttm,nouveau
crc32_pclmul 16384 0
video 40960 2 asus_wmi,nouveau
crc32c_intel 20480 0
ghash_clmulni_intel 16384 0
iTCO_wdt 16384 0
intel_cstate 16384 0
iTCO_vendor_support 16384 1 iTCO_wdt
e1000e 204800 0
wmi_bmof 16384 0
smartpqi 57344 2
intel_uncore 86016 0
mxm_wmi 16384 1 nouveau
scsi_transport_sas 28672 2 ses,smartpqi
mei_me 24576 0
joydev 24576 0
intel_rapl_perf 16384 0
ptp 20480 1 e1000e
pps_core 16384 1 ptp
pcc_cpufreq 16384 0
ioatdma 45056 0
mei 81920 1 mei_me
shpchp 32768 0
dca 16384 1 ioatdma
i2c_i801 24576 0
acpi_cpufreq 20480 0
wmi 20480 4 asus_wmi,wmi_bmof,mxm_wmi,nouveau
raid10 45056 0
raid456 139264 0
async_raid6_recov 16384 1 raid456
async_pq 16384 2 raid456,async_raid6_recov
async_xor 16384 3 async_pq,raid456,async_raid6_recov
async_memcpy 16384 2 raid456,async_raid6_recov
async_tx 16384 5 async_pq,async_memcpy,async_xor,raid456,async_raid6_recov
raid1 36864 0
raid0 20480 0
multipath 16384 0
linear 16384 0
nvme 36864 0
uas 20480 0
nvme_core 45056 1 nvme
usb_storage 69632 1 uas
|
note 1: i don't have /etc/udev/rules.d/70-net-name-use-custom.rules and i keep the system to do its job itself.
note 2: yes i've tried a different router port and the behavior was the same :-/ |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54237 Location: 56N 3W
|
Posted: Tue Apr 16, 2019 8:17 pm Post subject: |
|
|
JohnBlbec,
I'm looking for a self consistent data set from your Gentoo that's not working.
That's dmesg, the complete unstripped kernel configuration that produced the dmseg, the lspci -knn output which we have now.
Even if you only have one kernel built, it can still be the wrong kernel. Consider the following ...
You don't mount boot but you build and install the kernel anyway. The kernel binary correctly installs to the wrong place, so the boot loader never sees it.
The kernel modules correctly install to the right place, overwriting the modules that were there before.
When you reboot, you have the old kernel binary and the new modules. This sometimes sort of works.
Checking your is part of the process of ensuring that you are providing the self consistent data set you think you are.
Many users come here with problem that they have actually already fixed but they don't know it because they made a mess of the kernel install and they are not running the kernel they think they are.
Code: | 00:1f.6 Ethernet controller [0200]: Intel Corporation Ethernet Connection (2) I219-V [8086:15b8]
Subsystem: ASUSTeK Computer Inc. Ethernet Connection (2) I219-V [1043:8672]
Kernel driver in use: e1000e
Kernel modules: e1000e | Taht all looks good.
e1000e and e1000 are fairly similar. Indeed, I've seen cards move between then for one kernel to the next, so to be on the safe side, build both e1000e and e1000. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Tue Apr 16, 2019 9:14 pm Post subject: |
|
|
additional information - booted not working system
Code: |
# uname -a
Linux rpc-linux-new 4.19.27-gentoo-r1 #1 SMP Sun Mar 24 22:41:05 CET 2019 x86_64 Intel(R) Core(TM) i9-7940X CPU @ 3.10GHz GenuineIntel GNU/Linux
|
Code: |
# ls -la /lib/modules
total 20
drwxr-xr-x 3 root root 4096 Mar 24 22:57 .
drwxr-xr-x 16 root root 12288 Mar 24 23:40 ..
drwxr-xr-x 5 root root 4096 Mar 24 22:43 4.19.27-gentoo-r1
|
Code: |
# ls -la /boot
total 572640
drwxr-xr-x 4 root root 4096 Jan 1 1970 .
drwxr-xr-x 21 root root 4096 Mar 24 22:59 ..
drwxr-xr-x 3 root root 4096 Mar 2 23:46 EFI
-rwxr-xr-x 1 root root 2719580 Mar 24 21:41 System.map-genkernel-x86_64-4.19.27-gentoo-r1
drwxr-xr-x 6 root root 4096 Mar 24 21:57 grub
-rwxr-xr-x 1 root root 4040644 Mar 24 21:44 initramfs-genkernel-x86_64-4.19.27-gentoo-r1
-rwxr-xr-x 1 root root 5685632 Mar 24 21:41 kernel-genkernel-x86_64-4.19.27-gentoo-r1
-rwxr-xr-x 1 root root 573913088 Oct 26 05:21 systemrescuecd-x86-5.3.1.iso
|
for building my kernel i always use my script which is working well for years. kernel's been compiled without error and also installation's been successfull.
Code: |
#!/bin/bash
#
# $Id: make_kernel.sh,v 1.8 2019/03/11 10:43:06 root Exp $
#
config="$1"
if [ -z "$config" -o "$config" = "-h" -o "$config" = "--help" ]; then
echo "use: $(basename 0) <kernel_config>"
exit 1
fi
if [ ! -f "$config" ]; then
echo "error: '$config' doesn't exist"
exit 2
fi
is_tmpfs_mounted()
{
df -P | grep -q '/var/tmpfs' &> /dev/null
}
is_emerge_in_progress()
{
ps aux | grep -q [e]merge &> /dev/null
}
trap "{ is_tmpfs_mounted && is_emerge_in_progress && umount /var/tmpfs; }" SIGINT SIGTERM
is_tmpfs_mounted || mount /var/tmpfs
genkernel --menuconfig --save-config --mountboot --install --lvm --e2fsprogs --no-splash --disklabel --makeopts="-j28" --tempdir=/var/tmpfs/genkernel --kernel-config="$config" all
[[ $? -eq 0 ]] && grub-mkconfig -o /boot/grub/grub.cfg
is_tmpfs_mounted && is_emerge_in_progress && umount /var/tmpfs
|
here is an output of lsmod
here is a complete dmesg output
here is a complete kernel config kernel-config-x86_64-4.19.27-gentoo-r1
i have also trouble to umount usb just mounted stick with exfat - umount stuck and it is not possible to break it using ctrl+c. "kill -9" also doesn't work. btw. stick with fat32 is ok. as i said "/etc/init.d/net.enp0s31f6 stop" also doesn't work and stuck the same way as umount. very similar problem i have during shutdown - it sometimes stuck during umount /var, sometimes it stuck farther during shuting down lvm but never finish with success. i've checked my /etc/fstab but it looks ok by my opinion. those things were already working in some of my previous kernel config so i ignore them for now. i've mentioned it because maybe i overlook something and you'll be able to see possible dependencies...
here is my /etc/fstab |
|
Back to top |
|
|
NeddySeagoon Administrator
Joined: 05 Jul 2003 Posts: 54237 Location: 56N 3W
|
Posted: Tue Apr 16, 2019 10:52 pm Post subject: |
|
|
JohnBlbec,
Your kernel is timestamed Mar 24 21:41 in /boot, which ties in with Mar 24 22:43 for /lib/modules/4.19.27-gentoo-r1
Thats about 3 weeks ago. Its also consistent with Sun Mar 24 22:41:05 in your uname -a.
Good. I'm surprised by the 3 week old kernel but its far more important that the bits are consistent.
That ties in with the Sun Mar 24 22:41:05 at the top of dmesg.
dmesg looks good to Code: | [ 10.553381] IPv6: ADDRCONF(NETDEV_UP): enp0s31f6: link is not ready |
I suspect that you need one of the Code: | MII PHY device drivers | but I don't know which one. _________________ Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Tue Apr 16, 2019 11:20 pm Post subject: |
|
|
sometimes a man is too tired trying to solve his problem and it is better to let the problem sleep. that's the case. I stopped to solve kernel and started to find out a problem somewhere else. as you can see I've failed that's why 3 weeks old kernel. frankly speaking I was desperate...
tomorrow I'll try to think about MII PHY device driver and recompile the kernel. thanks for the hint! I'll keep you informed. |
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Wed Apr 17, 2019 4:55 pm Post subject: |
|
|
i've added MII below but is hasn't helped :-(
Code: |
CONFIG_DMIID=y
CONFIG_MII=m
|
the other MII in the kernel are not for intel...
[edited]
i've added also one (see below) because systemrescuecd has it. it hasn't helped.
Code: |
CONFIG_XILINX_GMII2RGMII=m
|
what should i change if i could try systemrescuecd kernel config to be bootable on my pc? i've tried it several times but it's never booted. i have no experience with livecd and i have just now no other idea :-(
Last edited by JohnBlbec on Wed Apr 17, 2019 5:28 pm; edited 1 time in total |
|
Back to top |
|
|
Tony0945 Watchman
Joined: 25 Jul 2006 Posts: 5127 Location: Illinois, USA
|
Posted: Wed Apr 17, 2019 5:08 pm Post subject: |
|
|
If I may interject a suggestion:
Build a generic genkernel kernel by emerging genkernel and running it. Temporarily edit your boot loader to load it instead of your custom kernel.
Boot. if the problem persists it is 99% NOT the kernel. If the problem is fixed, it was the kernel and you should compare the configs. |
|
Back to top |
|
|
JohnBlbec Guru
Joined: 08 Feb 2003 Posts: 306
|
Posted: Wed Apr 17, 2019 5:30 pm Post subject: |
|
|
Tony0945 wrote: | If I may interject a suggestion:
Build a generic genkernel kernel by emerging genkernel and running it. Temporarily edit your boot loader to load it instead of your custom kernel.
Boot. if the problem persists it is 99% NOT the kernel. If the problem is fixed, it was the kernel and you should compare the configs. |
ok but i guess there will have to be done some changes. is uefi boot supported in the generic genkernel config for example? |
|
Back to top |
|
|
|