You are not logged in.
I have just upgraded my generic amd64 box, because
- it was freezing on X screenblank since upgrading to 4.15.1-2-ARCH.
- I hoped the upgrade to 4.15.2-2-ARCH might solve the problem which I assumed to be related to amdgpu
However the upgrade throws a different error
...
:: Exécution des crochets de pré-transaction...
(1/1) Remove DKMS modules
==> dkms remove r8169_aspm/v1.1 -k 4.15.1-2-ARCH
Error! There is no instance of r8169_aspm v1.1
for kernel 4.15.1-2-ARCH (x86_64) located in the DKMS tree.
...
...
:: Exécution des crochets de post-transaction...
( 1/11) Updating linux module dependencies...
( 2/11) Install DKMS modules
==> dkms install r8169_aspm/v1.1 -k 4.15.2-2-ARCH
Error! Bad return status for module build on kernel: 4.15.2-2-ARCH (x86_64)
Consult /var/lib/dkms/r8169_aspm/v1.1/build/make.log for more information.
( 3/11) Updating linux initcpios...
...
and /var/lib/dkms/r8169_aspm/v1.1/build/make.log shows
DKMS make.log for r8169_aspm-v1.1 for kernel 4.15.2-2-ARCH (x86_64)
sam. févr. 10 13:25:24 CET 2018
make : on entre dans le répertoire « /usr/lib/modules/4.15.2-2-ARCH/build »
CC [M] /var/lib/dkms/r8169_aspm/v1.1/build/r8169_aspm.o
/var/lib/dkms/r8169_aspm/v1.1/build/r8169_aspm.c: Dans la fonction « rtl_init_one »:
/var/lib/dkms/r8169_aspm/v1.1/build/r8169_aspm.c:8536:2: error: déclaration implicite de la fonction « setup_timer »; vouliez-vous utiliser « sk_stop_timer » ? [-Werror=implicit-function-declaration]
setup_timer(&tp->timer, rtl8169_phy_timer, (unsigned long)dev);
^~~~~~~~~~~
sk_stop_timer
cc1 : certains avertissements sont traités comme des erreurs
make[1]: *** [scripts/Makefile.build:323: /var/lib/dkms/r8169_aspm/v1.1/build/r8169_aspm.o] Error 1
make: *** [Makefile:1508: _module_/var/lib/dkms/r8169_aspm/v1.1/build] Error 2
make : on quitte le répertoire « /usr/lib/modules/4.15.2-2-ARCH/build »
I am out of my depth here. What should I do about the realtek module r8169 issue?
(I may need to open a separate thread about the possible amdgpu system freeze problem.)
TIA
Offline
Similar problem with rtbth-dkms module for RT3290 bluetooth support after updating to kernel 4.15.2-2-ARCH
DKMS make.log for rtbth-3.9.4.4_8aa074e for kernel 4.15.2-2-ARCH (x86_64)
Sun Feb 11 00:34:59 CET 2018
make: Entering directory '/usr/lib/modules/4.15.2-2-ARCH/build'
CC [M] /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_core_main.o
CC [M] /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_core_init.o
CC [M] /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_core_pci.o
CC [M] /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_core_bluez.o
CC [M] /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_core_us.o
CC [M] /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_hlpr_hw.o
CC [M] /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_hlpr_dbg.o
CC [M] /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_hlpr_linux.o
/var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_hlpr_linux.c: In function ‘ral_timer_init’:
/var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_hlpr_linux.c:428:3: error: implicit declaration of function ‘setup_timer’; did you mean ‘setup_irq’? [-Werror=implicit-function-declaration]
setup_timer(timer, (void *)func, (unsigned long)os_timer);
^~~~~~~~~~~
setup_irq
cc1: some warnings being treated as errors
make[1]: *** [scripts/Makefile.build:316: /var/lib/dkms/rtbth/3.9.4.4_8aa074e/build/rtbth_hlpr_linux.o] Error 1
make[1]: *** Waiting for unfinished jobs....
make: *** [Makefile:1508: _module_/var/lib/dkms/rtbth/3.9.4.4_8aa074e/build] Error 2
make: Leaving directory '/usr/lib/modules/4.15.2-2-ARCH/build'
Will update this post if i find relevant workaround for this issue.
Offline
These are AUR packages; it is your job, not pacman's, to update them:
https://wiki.archlinux.org/index.php/Ar … g_packages
Moving to AUR Issues.
Offline
For r8169_aspm
diff --git a/r8169_aspm.c b/r8169_aspm.c
index 41a51c7..3b4dee8 100644
--- a/r8169_aspm.c
+++ b/r8169_aspm.c
@@ -4405,10 +4405,9 @@ static void rtl_schedule_task(struct rtl8169_private *tp, enum rtl_flag flag)
schedule_work(&tp->wk.work);
}
-static void rtl8169_phy_timer(unsigned long __opaque)
+static void rtl8169_phy_timer(struct timer_list *t)
{
- struct net_device *dev = (struct net_device *)__opaque;
- struct rtl8169_private *tp = netdev_priv(dev);
+ struct rtl8169_private *tp = from_timer(tp, t, timer);
rtl_schedule_task(tp, RTL_FLAG_TASK_PHY_PENDING);
}
@@ -8533,7 +8532,7 @@ static int rtl_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
tp->opts1_mask = (tp->mac_version != RTL_GIGA_MAC_VER_01) ?
~(RxBOVF | RxFOVF) : ~0;
- setup_timer(&tp->timer, rtl8169_phy_timer, (unsigned long)dev);
+ timer_setup(&tp->timer, rtl8169_phy_timer, 0);
tp->rtl_fw = RTL_FIRMWARE_UNKNOWN;
For rtbth-dkms source on github has been deleted and maintainer has not addressed the issue in over a year.
Edit:
For rtbth-dkms-git is the function ral_timer_init ever called? As very nasty test could put BUG() at the start of the function comment out the setup_timer call and see if the bug is ever reached.
Last edited by loqs (2018-02-11 17:57:13)
Offline