On Tue, 21 Feb 2017, Fengguang Wu wrote:
Hi Marc,
FYI here is another bisect result. The attached reproduce-* script can
be used to reproduce the bug.
Again. This is a problem in the calling code The WARN_ON merily shows the
wreckage in the caller. Marc sent you a patch already...
[ 11.409712] hrtimer_init+0x11f/0x199
[ 11.410666] ? mac80211_hwsim_get_tsf+0x1d/0x1d
[ 11.411766] tasklet_hrtimer_init+0x1b/0x4f
[ 11.412802] mac80211_hwsim_new_radio+0x7fe/0x916
[ 11.413935] ? set_debug_rodata+0x12/0x12
[ 11.414904] init_mac80211_hwsim+0x138/0x29f
[ 11.415822] ? rndis_wlan_driver_init+0x1b/0x1b
[ 11.416775] do_one_initcall+0x90/0x142
[ 11.417632] ? set_debug_rodata+0x12/0x12
[ 11.418511] kernel_init_freeable+0x1cb/0x258
[ 11.419433] ? rest_init+0x13b/0x13b
[ 11.420233] kernel_init+0xe/0xf5
[ 11.421006] ret_from_fork+0x2a/0x40
[ 11.421838] ---[ end trace 9c23eceab0d16aa5 ]---
[ 11.423540] ieee80211 phy1: Selected rate control algorithm 'minstrel_ht'