Bug#973921: Kernel trace: connmand tainted (PHY driver not loaded?)

[Camaleón <noelamac@gmail.com>]

Version: 5.10.0-3

El 2021-02-14 a las 09:25 +0100, Salvatore Bonaccorso escribió:

> On Sat, Nov 07, 2020 at 03:47:24PM +0100, Camaleón wrote:
> > Package: src:linux 
> > Version: 5.9.0-1
> > Severity: normal
> > 
> > Hello,
> > 
> > I'm getting this call trace on every boot up, with no 
> > apparent/noticeable side effects. I mean, wired and wireless network 
> > cards are both detected and working (full dmesg is attached).
> > 
> > Yes, «connman» package is installed and also, apparently working as it 
> > should. 
> > 
> > [   28.195123] ------------[ cut here ]------------
> > [   28.195127] read_page callback not available, PHY driver not loaded?
> > [   28.195250] WARNING: CPU: 0 PID: 366 at drivers/net/phy/phy-core.c:719 phy_save_page+0x5b/0x60 [libphy]
> > [   28.195253] Modules linked in: iptable_nat nf_nat nf_conntrack 
> > nf_defrag_ipv6 nf_defrag_ipv4 iptable_mangle iptable_filter brcmsmac 
> > brcmutil cordic uvcvideo mac80211 videobuf2_vmalloc videobuf2_memops 
> > videobuf2_v4l2 videobuf2_common videodev cfg80211 mc hp_wmi libarc4 
> > sparse_keymap r8169 snd_hda_codec_idt at24 snd_hda_codec_generic rfkill 
> > ledtrig_audio snd_hda_intel snd_intel_dspcfg realtek snd_hda_codec 
> > wmi_bmof mdio_devres joydev pcspkr coretemp rts5208(C) libphy 
> > snd_hda_core i2c_i801 snd_hwdep sg snd_pcm bcma uhci_hcd wmi ehci_pci 
> > i2c_smbus snd_timer ehci_hcd snd soundcore usbcore usb_common lpc_ich 
> > ac battery button acpi_cpufreq ip_tables x_tables autofs4 xfs libcrc32c 
> > crc32c_generic sd_mod t10_pi crc_t10dif crct10dif_generic 
> > crct10dif_common i915 ahci libahci libata scsi_mod i2c_algo_bit 
> > drm_kms_helper psmouse evdev cec drm serio_raw video
> > [   28.195366] CPU: 0 PID: 366 Comm: connmand Tainted: G         C        5.9.0-1-amd64 #1 Debian 5.9.1-1

(...)

> Unless mistaken his was introduce due to issues in the module
> subsystem preventing load of the PHY driver module on certain systems,
> as described in
> https://lore.kernel.org/netdev/20191018.125825.843031213764652652.davem@davemloft.net/t/
> which added the WARN_ONCE.
> 
> Not sure if the underlying issues were issues in meanwhile.

The warning is now gone (current kernel 5.10) so I guess this bug can 
be safely closed.

Thanks for noticing.

Cheers,

-- 
Camaleón