Source: linux Version: 6.19.8-1 Severity: normal Tags: upstream X-Debbugs-Cc: [email protected], [email protected] User: [email protected] Usertags: loong64
Hi, DSA reinstalled a loong64 physical machine (previously using the debian-ports archive, now running latest sid), and we started observing kernel warnings when starting a VM using KVM. First warning: | [ 2050.507635] ------------[ cut here ]------------ | [ 2050.507662] memcpy: detected field-spanning write (size 4) of single field "p" at arch/loongarch/kvm/intc/eiointc.c:520 (size 0) | [ 2050.507682] WARNING: arch/loongarch/kvm/intc/eiointc.c:520 at kvm_eiointc_regs_access.isra.0+0x354/0x3c0, CPU#6: qemu-system-loo/16813 | [ 2050.507697] Modules linked in: bridge stp llc nls_ascii nls_cp437 vfat fat snd_hda_intel snd_intel_dspcfg snd_hda_codec snd_hda_core ast snd_hwdep drm_client_lib snd_pcm drm_shmem_helper sg drm_kms_helper snd_timer snd i2c_algo_bit evdev soundcore ip6t_REJECT nf_reject_ipv6 ip6table_filter ip6_tables xt_hashlimit ipt_REJECT nf_reject_ipv4 xt_NFLOG nfnetlink_log xt_multiport xt_tcpudp xt_conntrack nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 drm iptable_filter ip_tables x_tables dm_snapshot dm_bufio vhost_net vhost tun vhost_iotlb tap sch_fq tcp_bbr zlib_deflate configfs nfnetlink autofs4 ext4 crc16 mbcache jbd2 crc32c_cryptoapi raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq raid0 dm_mod raid1 md_mod sd_mod cdc_ether usbnet mii ahci libahci xhci_pci ohci_pci xhci_hcd libata ehci_pci dwmac_loongson ehci_hcd stmmac_libpci megaraid_sas ohci_hcd stmmac usbcore scsi_mod r8169 pcs_xpcs realtek phylink scsi_common usb_common efivarfs | [ 2050.507873] CPU: 6 UID: 0 PID: 16813 Comm: qemu-system-loo Not tainted 6.19.8+deb14-loong64 #1 PREEMPTLAZY Debian 6.19.8-1 | [ 2050.507879] Hardware name: LOONGSON Dabieshan/Loongson-LS2C50C6, BIOS Loongson UEFI (3C50007A2000_C6) V4.3.0-Dual 05/21/25 09:17:40 | [ 2050.507883] pc 9000000000280d74 ra 9000000000280d74 tp 900000010faac000 sp 900000010faafb20 | [ 2050.507887] a0 0000000000000074 a1 0000000000000000 a2 900000010faaf920 a3 900000010faaf918 | [ 2050.507890] a4 0000000000000000 a5 9000000001a8d960 a6 203a7970636d656d a7 293020657a697328 | [ 2050.507894] t0 6d08ee86308d20d5 t1 6d08ee86308d20d5 t2 90000000017e8000 t3 0000000000000001 | [ 2050.507897] t4 fffffffffffffffe t5 00000000ffffdfff t6 900010207ff04000 t7 0000000000000000 | [ 2050.507900] t8 0000000000000000 u0 900000011ad00050 s9 900000010faafec0 s0 900000010faafb90 | [ 2050.507903] s1 900000011ad00000 s2 900000011ad00050 s3 0000000000000000 s4 00007ffffbf39830 | [ 2050.507906] s5 000000000000002f s6 000055556b5480d0 s7 0000555559d0be48 s8 0000000000000000 | [ 2050.507910] ra: 9000000000280d74 kvm_eiointc_regs_access.isra.0+0x354/0x3c0 | [ 2050.507914] ERA: 9000000000280d74 kvm_eiointc_regs_access.isra.0+0x354/0x3c0 | [ 2050.507918] CRMD: 000000b0 (PLV0 -IE -DA +PG DACF=CC DACM=CC -WE) | [ 2050.507932] PRMD: 00000000 (PPLV0 -PIE -PWE) | [ 2050.507940] EUEN: 00000007 (+FPE +SXE +ASXE -BTE) | [ 2050.507948] ECFG: 00071c1d (LIE=0,2-4,10-12 VS=7) | [ 2050.507956] ESTAT: 000c0000 [BRK] (IS= ECode=12 EsubCode=0) | [ 2050.507962] PRID: 0014c011 (Loongson-64bit, Loongson-3C5000) | [ 2050.507966] CPU: 6 UID: 0 PID: 16813 Comm: qemu-system-loo Not tainted 6.19.8+deb14-loong64 #1 PREEMPTLAZY Debian 6.19.8-1 | [ 2050.507970] Hardware name: LOONGSON Dabieshan/Loongson-LS2C50C6, BIOS Loongson UEFI (3C50007A2000_C6) V4.3.0-Dual 05/21/25 09:17:40 | [ 2050.507972] Stack : 900000010faaf7f8 0000000000000000 9000000000238828 900000010faac000 | [ 2050.507978] 900000010faaf740 900000010faaf748 0000000000000000 900000010faaf888 | [ 2050.507983] 900000010faaf880 900000010faaf880 900010207ff19b40 6572617764726148 | [ 2050.507987] 203a656d616e2065 900000010faaf748 6d08ee86308d20d5 900000012859c840 | [ 2050.507992] 900000010faac000 90000000015c8868 00000000ffffdfff 900010207ff04000 | [ 2050.507997] 0000000000000000 0000000000000000 000000207b878000 900000010faafec0 | [ 2050.508002] 0000000000000000 90000000017e8000 0000000000000000 90000000015c8868 | [ 2050.508006] 0000000000000208 0000000000000009 000055556b5480d0 0000555559d0be48 | [ 2050.508011] 0000000000000000 0000000000000000 9000000000238844 000055556bcb67e8 | [ 2050.508016] 00000000000000b0 0000000000000000 0000000000000007 0000000000071c1d | [ 2050.508020] ... | [ 2050.508023] Call Trace: | [ 2050.508026] [<9000000000238844>] show_stack+0x64/0x190 | [ 2050.508037] [<9000000000230fc8>] dump_stack_lvl+0x70/0x9c | [ 2050.508041] [<9000000000289630>] __warn+0xa0/0x1b0 | [ 2050.508046] [<90000000012358e8>] __report_bug+0xa8/0x1c0 | [ 2050.508052] [<9000000001235af0>] report_bug+0x40/0xd0 | [ 2050.508055] [<90000000012791f4>] do_bp+0x254/0x420 | [ 2050.508066] [<0000000000000000>] 0x0 | [ 2050.508070] [<9000000000280d74>] kvm_eiointc_regs_access.isra.0+0x354/0x3c0 | [ 2050.508073] [<90000000002812bc>] kvm_eiointc_set_attr+0x34c/0x770 | [ 2050.508076] [<9000000000262f04>] kvm_device_ioctl+0x264/0x3a0 | [ 2050.508082] [<900000000075f09c>] sys_ioctl+0x52c/0x1150 | [ 2050.508089] [<9000000001279804>] do_syscall+0xc4/0x320 | [ 2050.508094] ---[ end trace 0000000000000000 ]--- This warning could have been introduced by the following upstream commit: commit 01a8e68396a6d51f5ba92021ad1a4b8eaabdd0e7 Author: Bibo Mao <[email protected]> Date: Thu Sep 18 19:44:22 2025 +0800 LoongArch: KVM: Avoid copy_*_user() with lock hold in kvm_eiointc_sw_status_access() The second one: | [ 2050.508176] ------------[ cut here ]------------ | [ 2050.508179] UBSAN: array-index-out-of-bounds in /build/reproducible-path/linux-6.19.8/arch/loongarch/kvm/vcpu.c:569:20 | [ 2050.508234] index -1 is out of range for type 'kvm_phyid_info [256]' | [ 2050.508248] CPU: 6 UID: 0 PID: 16813 Comm: qemu-system-loo Tainted: G W 6.19.8+deb14-loong64 #1 PREEMPTLAZY Debian 6.19.8-1 | [ 2050.508253] Tainted: [W]=WARN | [ 2050.508254] Hardware name: LOONGSON Dabieshan/Loongson-LS2C50C6, BIOS Loongson UEFI (3C50007A2000_C6) V4.3.0-Dual 05/21/25 09:17:40 | [ 2050.508256] Stack : 900000010faaf8f8 0000000000000000 9000000000238828 900000010faac000 | [ 2050.508261] 900000010faaf840 900000010faaf848 0000000000000000 900000010faaf988 | [ 2050.508266] 900000010faaf980 900000010faaf980 900010207ff1a3f0 6572617764726148 | [ 2050.508271] 203a656d616e2065 900000010faaf848 6d08ee86308d20d5 900000012859c840 | [ 2050.508275] 900000010faac000 90000000015c8868 00000000ffffdfff 900010207ff04000 | [ 2050.508280] 0000000000000000 0000000000000000 000000207b878000 0000000000000000 | [ 2050.508284] 0000000000000000 90000000017e8000 0000000000000000 90000000015c8868 | [ 2050.508289] 0000000000000003 ffffffffffffffff 900000011ad020b8 0000000000000000 | [ 2050.508293] 900000011ad00000 0000000000000000 9000000000238844 000055556bcb67e8 | [ 2050.508298] 00000000000000b0 0000000000000007 0000000000000007 0000000000071c1d | [ 2050.508302] ... | [ 2050.508304] Call Trace: | [ 2050.508306] [<9000000000238844>] show_stack+0x64/0x190 | [ 2050.508310] [<9000000000230fc8>] dump_stack_lvl+0x70/0x9c | [ 2050.508314] [<900000000022bc60>] ubsan_epilogue+0xc/0x3c | [ 2050.508318] [<9000000000bc61c4>] __ubsan_handle_out_of_bounds+0xa4/0xb0 | [ 2050.508322] [<900000000027a6c0>] kvm_get_vcpu_by_cpuid+0xb0/0xc0 | [ 2050.508326] [<90000000002813b8>] kvm_eiointc_set_attr+0x448/0x770 | [ 2050.508329] [<9000000000262f04>] kvm_device_ioctl+0x264/0x3a0 | [ 2050.508334] [<900000000075f09c>] sys_ioctl+0x52c/0x1150 | [ 2050.508338] [<9000000001279804>] do_syscall+0xc4/0x320 | [ 2050.508343] ---[ end trace ]--- The problem was not present when running kernel 6.17.7+deb14-loong64, so it seems to be a relatively recent regression. Despite the warnings, things seems to work relatively well. Regards Aurelien
