No GPIO on armv7 Allwinner A20; the /sys/class/gpio directory is missing. OK on 4.8.17 kernel. This makes the kernel unusable in embedded applications.
Changing my +1 to -1. Regression in e1000e module. My I218-LM adapter in T450s is having trouble reconnecting after suspend/resume. WiFi works fine connecting to the same router. Also, a cold reboot makes things work again.
Reverting to 4.8.16 fixes the e1000e suspend/resume issue. Upon resume, ethtool reports speed as 1000Mb/s, rather than 10Mb/s that is reported with 4.9.2 kernel on resume.
@dimitrisk: Thanks. I looked at recent commits to e1000e in the kernel tree and nothing really stands out, but who knows. Also not sure if something else may be causing the difference here when it comes to X250 v. T450s regarding this adapter. I filed bugs in Fedora/Kernel bugzillas. We'll see what happens.
@funfunctor: I have linux-firmware-20160923-68.git42ad5367.fc25 installed on this machine, which appears to be from stable. So, not quite sure how to answer this, but I'm guessing no.
Just tried 4.9.3 from koji and got pretty much the same result with suspend/resume on T450s. The first suspend/resume cycle was OK this time, but the next one broke ethernet.
Just tried it this morning, get a blank screen (monitors in power saving mode after grub attempts to boot). Booting an older kernel works fine. Will do more research and open a bz.
This update has been submitted for testing by labbott.
Looks good to me
No GPIO on armv7 Allwinner A20; the /sys/class/gpio directory is missing. OK on 4.8.17 kernel. This makes the kernel unusable in embedded applications.
No regressions noted on x86_64 Ivy Bridge and i686 Diamondville.
Works for me, X86_64. No regressions noticed.
wfm - desktop 32GB Intel i7-4770 CPU and laptop 16GB Intel i7-3610QM CPU
Took over 80 minutes to complete the performance tests on the laptop!!!
Warnings at start of boot. CIFS shares won't stay mounted. Was fine on 4.8.16.
This update has been pushed to testing.
Boots on ThinkPad T450s, an Atom based server and a VM, all x86_64.
No regression noted, XPS 15 9550/0N7TVV
Looks good
Changing my +1 to -1. Regression in e1000e module. My I218-LM adapter in T450s is having trouble reconnecting after suspend/resume. WiFi works fine connecting to the same router. Also, a cold reboot makes things work again.
Reverting to 4.8.16 fixes the e1000e suspend/resume issue. Upon resume, ethtool reports speed as 1000Mb/s, rather than 10Mb/s that is reported with 4.9.2 kernel on resume.
works for me on t450s
@cserpentis: and e1000e driver works for you after suspend/resume?
Works for me.
Works great on Lenovo X220! LGTM! =)
Works well on my fedora 25 x86_64 virtual box 5.1.12
karma: +1
Thinkpad X250, WFM.
@bojan: I have the same Ethernet adapter, works for me across suspend/resume, speed stays at 1000Mbps according to ethtool.
@dimitrisk: Thanks. I looked at recent commits to e1000e in the kernel tree and nothing really stands out, but who knows. Also not sure if something else may be causing the difference here when it comes to X250 v. T450s regarding this adapter. I filed bugs in Fedora/Kernel bugzillas. We'll see what happens.
@bojan did linux-firmware get updated and mess up for you?
@funfunctor: I have linux-firmware-20160923-68.git42ad5367.fc25 installed on this machine, which appears to be from stable. So, not quite sure how to answer this, but I'm guessing no.
LGTM , pass default and performance test
LGTM
my laptop boot fine, no issues noted
Bug 1412863 has been opened for the missing GPIO sysfs.
no regressions noted here i7, network connections are fine
Just tried 4.9.3 from koji and got pretty much the same result with suspend/resume on T450s. The first suspend/resume cycle was OK this time, but the next one broke ethernet.
WFM, Lenovo Flex4, no regressions, wifi and basic performance checks are functioning as expected.
WGFM as noted in bodhi comment
Everything ok on T440s
This update has been obsoleted by kernel-4.9.3-200.fc25.
Just tried it this morning, get a blank screen (monitors in power saving mode after grub attempts to boot). Booting an older kernel works fine. Will do more research and open a bz.
Sorry, I fail, checked +1, this should have been a -1.