On Tue, Aug 6, 2019 at 5:46 PM Qian Cai <cai@xxxxxx> wrote: > > It looks like the linux-next commit “of/platform: Pause/resume sync state during init and of_platform_populate()” [1] > Introduced a warning while booting arm64. > > [1] https://lore.kernel.org/lkml/20190731221721.187713-6-saravanak@xxxxxxxxxx/ > > [ 93.449300][ T1] arm-smmu-v3 arm-smmu-v3.4.auto: ias 44-bit, oas 44-bit (features 0x0000170d) > [ 93.464873][ T1] arm-smmu-v3 arm-smmu-v3.4.auto: allocated 524288 entries for cmdq > [ 93.485481][ T1] arm-smmu-v3 arm-smmu-v3.4.auto: allocated 524288 entries for evtq > [ 93.496320][ T1] arm-smmu-v3 arm-smmu-v3.5.auto: option mask 0x2 > [ 93.502917][ T1] arm-smmu-v3 arm-smmu-v3.5.auto: ias 44-bit, oas 44-bit (features 0x0000170d) > [ 93.621818][ T1] arm-smmu-v3 arm-smmu-v3.5.auto: allocated 524288 entries for cmdq > [ 93.643000][ T1] arm-smmu-v3 arm-smmu-v3.5.auto: allocated 524288 entries for evtq > [ 94.519445][ T1] libphy: Fixed MDIO Bus: probed > [ 94.524649][ T1] EFI Variables Facility v0.08 2004-May-17 > [ 94.601166][ T1] NET: Registered protocol family 17 > [ 94.766008][ T1] zswap: loaded using pool lzo/zbud > [ 94.774745][ T1] kmemleak: Kernel memory leak detector initialized (mempool size: 16384) > [ 94.774756][ T1699] kmemleak: Automatic memory scanning thread started > [ 94.812338][ T1368] pcieport 0000:0f:00.0: Adding to iommu group 0 > [ 94.984466][ T1] ------------[ cut here ]------------ > [ 94.989827][ T1] Unmatched sync_state pause/resume! > [ 94.989894][ T1] WARNING: CPU: 25 PID: 1 at drivers/base/core.c:691 device_links_supplier_sync_state_resume+0x100/0x128 > [ 95.006062][ T1] Modules linked in: > [ 95.009815][ T1] CPU: 25 PID: 1 Comm: swapper/0 Not tainted 5.3.0-rc3-next-20190806+ #11 > [ 95.018161][ T1] Hardware name: HPE Apollo 70 /C01_APACHE_MB , BIOS L50_5.13_1.11 06/18/2019 > [ 95.028593][ T1] pstate: 60400009 (nZCv daif +PAN -UAO) > [ 95.034077][ T1] pc : device_links_supplier_sync_state_resume+0x100/0x128 > [ 95.041124][ T1] lr : device_links_supplier_sync_state_resume+0x100/0x128 > [ 95.048167][ T1] sp : 34ff800806e6fbc0 > [ 95.052172][ T1] x29: 34ff800806e6fc00 x28: 0000000000000000 > [ 95.058177][ T1] x27: 0000000000000000 x26: 0000000000000000 > [ 95.064181][ T1] x25: 0000000000000038 x24: 0000000000000000 > [ 95.070185][ T1] x23: 0000000000000000 x22: 0000000000000019 > [ 95.076189][ T1] x21: 0000000000000000 x20: f9ff808b804e6c50 > [ 95.082193][ T1] x19: ffff100014a6e600 x18: 0000000000000040 > [ 95.088197][ T1] x17: 0000000000000000 x16: 86ff80099d581b50 > [ 95.094201][ T1] x15: 0000000000000000 x14: ffff100010086d1c > [ 95.100205][ T1] x13: ffff1000109d8688 x12: ffffffffffffffff > [ 95.106209][ T1] x11: 00000000000000f9 x10: ffff0808b804e6c6 > [ 95.112213][ T1] x9 : 4b71ad522c851d00 x8 : 4b71ad522c851d00 > [ 95.118217][ T1] x7 : 6170206574617473 x6 : ffff100014076972 > [ 95.124221][ T1] x5 : 34ff800806e6f8f0 x4 : 000000000000000f > [ 95.130225][ T1] x3 : ffff1000101bfa5c x2 : 0000000000000001 > [ 95.136229][ T1] x1 : 0000000000000001 x0 : 0000000000000022 > [ 95.142233][ T1] Call trace: > [ 95.145374][ T1] device_links_supplier_sync_state_resume+0x100/0x128 > [ 95.152074][ T1] of_platform_sync_state_init+0x10/0x1c > [ 95.157557][ T1] do_one_initcall+0x2f8/0x600 > [ 95.162172][ T1] do_initcall_level+0x37c/0x3fc > [ 95.166959][ T1] do_basic_setup+0x34/0x4c > [ 95.171313][ T1] kernel_init_freeable+0x188/0x24c > [ 95.176363][ T1] kernel_init+0x18/0x334 > [ 95.180543][ T1] ret_from_fork+0x10/0x18 > [ 95.184809][ T1] ---[ end trace a9ea68c902540fe5 ]--- > [ 95.269085][ T1] Freeing unused kernel memory: 28672K > [ 101.069860][ T1] Checked W+X mappings: passed, no W+X pages found > [ 101.076265][ T1] Run /init as init process > [ 101.186359][ T1] systemd[1]: System time before build time, advancing clock. I tested it again on my device (on an older kernel) and I don't see this warning. Is this on an ARM64 target without a populated DT? That's the only thing I can see that could cause this warning. This is literally the code with the matching pause/resume. I can't think of any other way the pause/resume could have ended up not matching. static int __init of_platform_default_populate_init(void) { struct device_node *node; if (!of_have_populated_dt()) return -ENODEV; platform_bus_type.add_links = of_link_to_suppliers; device_links_supplier_sync_state_pause(); <=========== PAUSE /* * Handle certain compatibles explicitly, since we don't want to create * platform_devices for every node in /reserved-memory with a * "compatible", */ for_each_matching_node(node, reserved_mem_matches) of_platform_device_create(node, NULL, NULL); node = of_find_node_by_path("/firmware"); if (node) { of_platform_populate(node, NULL, NULL, NULL); of_node_put(node); } /* Populate everything else. */ of_platform_default_populate(NULL, NULL, NULL); return 0; } arch_initcall_sync(of_platform_default_populate_init); static int __init of_platform_sync_state_init(void) { device_links_supplier_sync_state_resume(); <========= RESUME return 0; } late_initcall_sync(of_platform_sync_state_init); Thanks, Saravana
