Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
Ben Hutchings wrote... > As a temporary workaround, disabling the Spectre v2 mitigation with the > kernel parameter "nospectre_v2" should allow this kernel version to run > on older CPUs without SSE2. We'll fix this properly in a later update. Thank you, that did the right thing on my very old Celeron (Coppermine) CPU from around the year 2000, part of my hardware museum. Christoph signature.asc Description: PGP signature
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
On Fri, Aug 19, 2022 at 12:32 PM Etienne Vogt wrote: > > On Thu, 18 Aug 2022, Ben Hutchings wrote: > > >> I suspect someone thought it would be a good idea to compile the kernel > >> for P4 only, as both PIII and Athlon XP processors lack the SSE2 > >> instruction set. > > > > That was a good guess, though we don't change the configuration like > > that in stable updates. > > Well, that happened for Firefox, it became SSE2 only when they switched > to a higher upstream version in a stable update. 78.15.0esr-1~deb11u1 > is the last version that works on non SSE2 CPUs. Firefox is an entirely different issue. It uses Rust and, despite Debian's Rust maintainters' best intentions, upstream insists on making its 686 target use all the bells and whistles. Debian tried to mitigate this by generating code for plain old Pentium on i386, but it hasn't worked in recent times, because upstream has started treating all x86-32 targets as "just a number" instead of an actual CPU sepc. This also affects, among other things SVG-lib, which also migrated to Rust a few years ago. Martin-Éric
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
On Thu, 18 Aug 2022, Ben Hutchings wrote: I suspect someone thought it would be a good idea to compile the kernel for P4 only, as both PIII and Athlon XP processors lack the SSE2 instruction set. That was a good guess, though we don't change the configuration like that in stable updates. Well, that happened for Firefox, it became SSE2 only when they switched to a higher upstream version in a stable update. 78.15.0esr-1~deb11u1 is the last version that works on non SSE2 CPUs. The RETbleed mitigations, which are not needed on these CPUs or even functional on 32-bit kernels, interact with the Spectre v2 mitigations, which *are* used on these CPUs. And unfortunately the RETbleed mitigations added some unconditional LFENCE instructions, which should be conditional since they are part of SSE2. As a temporary workaround, disabling the Spectre v2 mitigation with the kernel parameter "nospectre_v2" should allow this kernel version to run on older CPUs without SSE2. We'll fix this properly in a later update. Glad to know that was just a mistake. I will stay with 5.10.0-16 until the next update shows up. -- Etienne Vogt (etienne.v...@obspm.fr) Observatoire de Paris-Meudon Service Informatique
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
On Thu, 2022-08-18 at 23:11 +0300, Martin-Éric Racine wrote: > On Thu, Aug 18, 2022 at 10:59 PM Ben Hutchings wrote: > > > > Control: retitle -1 [i386] Unconditional LFENCE instructions in > > FILL_RETURN_BUFFER > > Control: tag -1 confirmed upstream > > Control: found -1 5.18.14-1 > > > > On Wed, 2022-08-17 at 11:42 +0200, Etienne Vogt wrote: > > > I can confirm that this bug also occurs on Athlon XP systems (Generic VIA > > > KT333 motherboard, CPU AMD Athlon(tm) XP 2600+) : kernel panic early on > > > boot. > > > > > > I suspect someone thought it would be a good idea to compile the kernel > > > for P4 only, as both PIII and Athlon XP processors lack the SSE2 > > > instruction set. > > > > > > > That was a good guess, though we don't change the configuration like > > that in stable updates. > > > > The RETbleed mitigations, which are not needed on these CPUs or even > > functional on 32-bit kernels, interact with the Spectre v2 mitigations, > > which *are* used on these CPUs. And unfortunately the RETbleed > > mitigations added some unconditional LFENCE instructions, which should > > be conditional since they are part of SSE2. > > > > As a temporary workaround, disabling the Spectre v2 mitigation with the > > kernel parameter "nospectre_v2" should allow this kernel version to run > > on older CPUs without SSE2. We'll fix this properly in a later update. > > That breakage affects Stable. > > Expecting people to go and use workarounds on what was meant to be a > stable update isn't acceptable. Martin, you know we're all volunteers here, so don't take that tone. I was trying to be helpful in offering an alternative to holding back the upgrade. > I really hope that the fix will be pushed within the next 24 hours > with high urgency. This is unlikely to happen. Ben. -- Ben Hutchings I haven't lost my mind; it's backed up on tape somewhere. signature.asc Description: This is a digitally signed message part
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
On Thu, 2022-08-18 at 21:59 +0200, Ben Hutchings wrote: > Control: retitle -1 [i386] Unconditional LFENCE instructions in > FILL_RETURN_BUFFER > Control: tag -1 confirmed upstream > Control: found -1 5.18.14-1 > > On Wed, 2022-08-17 at 11:42 +0200, Etienne Vogt wrote: > > I can confirm that this bug also occurs on Athlon XP systems (Generic VIA > > KT333 motherboard, CPU AMD Athlon(tm) XP 2600+) : kernel panic early on > > boot. > > > > I suspect someone thought it would be a good idea to compile the kernel > > for P4 only, as both PIII and Athlon XP processors lack the SSE2 > > instruction set. > > > > That was a good guess, though we don't change the configuration like > that in stable updates. > > The RETbleed mitigations, which are not needed on these CPUs or even I mean PBRSB, not RETbleed. There are so many different issues to keep track of... Ben. > functional on 32-bit kernels, interact with the Spectre v2 mitigations, > which *are* used on these CPUs. And unfortunately the RETbleed > mitigations added some unconditional LFENCE instructions, which should > be conditional since they are part of SSE2. > > As a temporary workaround, disabling the Spectre v2 mitigation with the > kernel parameter "nospectre_v2" should allow this kernel version to run > on older CPUs without SSE2. We'll fix this properly in a later update. > > Ben. > -- Ben Hutchings I haven't lost my mind; it's backed up on tape somewhere. signature.asc Description: This is a digitally signed message part
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
On Thu, Aug 18, 2022 at 10:59 PM Ben Hutchings wrote: > > Control: retitle -1 [i386] Unconditional LFENCE instructions in > FILL_RETURN_BUFFER > Control: tag -1 confirmed upstream > Control: found -1 5.18.14-1 > > On Wed, 2022-08-17 at 11:42 +0200, Etienne Vogt wrote: > > I can confirm that this bug also occurs on Athlon XP systems (Generic VIA > > KT333 motherboard, CPU AMD Athlon(tm) XP 2600+) : kernel panic early on > > boot. > > > > I suspect someone thought it would be a good idea to compile the kernel > > for P4 only, as both PIII and Athlon XP processors lack the SSE2 > > instruction set. > > > > That was a good guess, though we don't change the configuration like > that in stable updates. > > The RETbleed mitigations, which are not needed on these CPUs or even > functional on 32-bit kernels, interact with the Spectre v2 mitigations, > which *are* used on these CPUs. And unfortunately the RETbleed > mitigations added some unconditional LFENCE instructions, which should > be conditional since they are part of SSE2. > > As a temporary workaround, disabling the Spectre v2 mitigation with the > kernel parameter "nospectre_v2" should allow this kernel version to run > on older CPUs without SSE2. We'll fix this properly in a later update. That breakage affects Stable. Expecting people to go and use workarounds on what was meant to be a stable update isn't acceptable. I really hope that the fix will be pushed within the next 24 hours with high urgency. Martin-Éric
Processed: Re: Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
Processing control commands: > retitle -1 [i386] Unconditional LFENCE instructions in FILL_RETURN_BUFFER Bug #1017425 [src:linux] 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII] Bug #1017494 [src:linux] linux-image-5.10.0-17-686-pae always attempts to kill idle task on Mobile P2 Changed Bug title to '[i386] Unconditional LFENCE instructions in FILL_RETURN_BUFFER' from '5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]'. Changed Bug title to '[i386] Unconditional LFENCE instructions in FILL_RETURN_BUFFER' from 'linux-image-5.10.0-17-686-pae always attempts to kill idle task on Mobile P2'. > tag -1 confirmed upstream Bug #1017425 [src:linux] [i386] Unconditional LFENCE instructions in FILL_RETURN_BUFFER Bug #1017494 [src:linux] [i386] Unconditional LFENCE instructions in FILL_RETURN_BUFFER Added tag(s) upstream and confirmed. Added tag(s) confirmed and upstream. > found -1 5.18.14-1 Bug #1017425 [src:linux] [i386] Unconditional LFENCE instructions in FILL_RETURN_BUFFER Bug #1017494 [src:linux] [i386] Unconditional LFENCE instructions in FILL_RETURN_BUFFER Marked as found in versions linux/5.18.14-1. Marked as found in versions linux/5.18.14-1. -- 1017425: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1017425 1017494: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1017494 Debian Bug Tracking System Contact ow...@bugs.debian.org with problems
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
Control: retitle -1 [i386] Unconditional LFENCE instructions in FILL_RETURN_BUFFER Control: tag -1 confirmed upstream Control: found -1 5.18.14-1 On Wed, 2022-08-17 at 11:42 +0200, Etienne Vogt wrote: > I can confirm that this bug also occurs on Athlon XP systems (Generic VIA > KT333 motherboard, CPU AMD Athlon(tm) XP 2600+) : kernel panic early on > boot. > > I suspect someone thought it would be a good idea to compile the kernel > for P4 only, as both PIII and Athlon XP processors lack the SSE2 > instruction set. > That was a good guess, though we don't change the configuration like that in stable updates. The RETbleed mitigations, which are not needed on these CPUs or even functional on 32-bit kernels, interact with the Spectre v2 mitigations, which *are* used on these CPUs. And unfortunately the RETbleed mitigations added some unconditional LFENCE instructions, which should be conditional since they are part of SSE2. As a temporary workaround, disabling the Spectre v2 mitigation with the kernel parameter "nospectre_v2" should allow this kernel version to run on older CPUs without SSE2. We'll fix this properly in a later update. Ben. -- Ben Hutchings I haven't lost my mind; it's backed up on tape somewhere. signature.asc Description: This is a digitally signed message part
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
I can confirm that this bug also occurs on Athlon XP systems (Generic VIA KT333 motherboard, CPU AMD Athlon(tm) XP 2600+) : kernel panic early on boot. I suspect someone thought it would be a good idea to compile the kernel for P4 only, as both PIII and Athlon XP processors lack the SSE2 instruction set. -- Etienne Vogt (etienne.v...@obspm.fr)
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
I have similar problem in Pentium III machine (DELL PowerEdge 1500SC). [Environment] CPU: Intel(R) Pentium(R) III CPU family (Tualatin) 1400MHz (SMP) Memory: 2GB Debian version: 11.4 linux-image-5.10.0-17-686-pae no longer boots correctly, while linux-image-5.10.0-16-686-pae works. In my case, the following kernel panic message is shown. [0.676076] Hardware name: Dell Computer Corporation PowerEdge 1500SC, BIOS A06 10/14/2004 [0.676134] EIP: __switch_to_asm+0x43/0x5c [0.676178] Code: 1d 54 45 ca c7 89 f6 bb 10 00 00 00 e8 07 00 00 00 f3 90 0f ae e8 eb f9 e8 07 00 00 00 f3 90 0f ae e8 eb f9 83 c4 08 4b 75 e2 <0f> ae e8 e8 01 00 00 00 cc 83 c4 04 0f ae e8 9d 5e 5f 5b 5d e9 78 [0.676238] EAX: c7a94480 EBX: ECX: 28494125 EDX: c114b340 [0.676280] ESI: f6fcea40 EDI: c7a94480 EBP: c7a8de3c ESP: c1145f9c [0.676322] DS: 007b ES: 007b FS: 00d8 CS: 00e0 SS: 0068 EFLAGS: 00210046 [0.676365] CR0: 80050033 CR2: ff9ff000 CR3: 07cb2000 CR4: 06f0 [0.676405] Call Trace: [0.676445] Modules linked in: [0.676503] ---[ end trace 02deed42763c6547 ]--- [0.676544] EIP: __switch_to_asm+0x43/0x5c [0.676583] Code: 1d 54 45 ca c7 89 f6 bb 10 00 00 00 e8 07 00 00 00 f3 90 0f ae e8 eb f9 e8 07 00 00 00 f3 90 0f ae e8 eb f9 83 c4 08 4b 75 e2 <0f> ae e8 e8 01 00 00 00 cc 83 c4 04 0f ae e8 9d 5e 5f 5b 5d e9 78 [0.676643) EAX: c7a94480 EBX: ECX: 28494125 EDX: c114b340 [0.676685] ESI: f6fcea40 EDI: c7a94480 EBP: c7a8de3c ESP: c1145f9c [0.676739] DS: 007b ES: 007b FS: 00d8 CS: 00e0 SS: 0068 EFLAGS: 00210046 [0.676793] CR0: 80050033 CR2: ff9ff000 CR3: 07cb2000 CR4: 06f0 [0.676847] Kernel panic - not syncing: Attempted to kill the idle task! [0.676989] ---[ end Kernel panic - not syncing: Attempted to kill the idle task! ]--- I hope this information helps. -- Takashi Yano
Bug#1017425: 5.10.0-17-686-pae: repeatedly crashes during initrd loading [PIII]
Package: src:linux Version: 5.10.136-1 Severity: grave -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 5.10.0-17-686-pae cannot boot on this Pentium III. The fan goes to full speed during initrd loading and the host reboots at that point. There is no screenshot to attach since there is no screen output after the "Loading initrd..." message from GRUB2. Reverting to 5.10.0-16-686-pae, the fan runs at low speed and the host boots normally. Martin-Éric - -- Package-specific info: processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 8 model name : Pentium III (Coppermine) stepping: 6 microcode : 0x8 cpu MHz : 865.440 cache size : 256 KB physical id : 0 siblings: 1 core id : 0 cpu cores : 1 apicid : 0 initial apicid : 0 fdiv_bug: no f00f_bug: no coma_bug: no fpu : yes fpu_exception : yes cpuid level : 2 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pse36 mmx fxsr sse cpuid pti bugs: cpu_meltdown spectre_v1 spectre_v2 spec_store_bypass l1tf mds swapgs itlb_multihit bogomips: 1730.88 clflush size: 32 cache_alignment : 32 address sizes : 36 bits physical, 32 bits virtual power management: ** Kernel log: boot messages should be attached ** Model information ** PCI devices: 00:00.0 Host bridge [0600]: Intel Corporation 82815 815 Chipset Host Bridge and Memory Controller Hub [8086:1130] (rev 02) Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- SERR- Kernel driver in use: agpgart-intel 00:01.0 PCI bridge [0604]: Intel Corporation 82815 815 Chipset AGP Bridge [8086:1131] (rev 02) (prog-if 00 [Normal decode]) Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap- 66MHz+ UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- SERR- TAbort- Reset- FastB2B- PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn- 00:1e.0 PCI bridge [0604]: Intel Corporation 82801 PCI Bridge [8086:244e] (rev 01) (prog-if 00 [Normal decode]) Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- SERR- TAbort- Reset- FastB2B- PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn- 00:1f.0 ISA bridge [0601]: Intel Corporation 82801BA ISA Bridge (LPC) [8086:2440] (rev 01) Control: I/O+ Mem+ BusMaster+ SpecCycle+ MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap- 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- TAbort- SERR- Kernel driver in use: nouveau Kernel modules: nouveau 02:04.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL8169 PCI Gigabit Ethernet Controller [10ec:8169] (rev 10) Subsystem: Realtek Semiconductor Co., Ltd. RTL8169/8110 Family PCI Gigabit Ethernet NIC [10ec:8169] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- SERR- Kernel driver in use: r8169 Kernel modules: r8169 02:08.0 Ethernet controller [0200]: Intel Corporation 82801BA/BAM/CA/CAM Ethernet Controller [8086:2449] (rev 01) Subsystem: Compaq Computer Corporation EtherExpress PRO/100 VM [0e11:0012] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- SERR- Kernel driver in use: e100 Kernel modules: e100 02:09.0 Ethernet controller [0200]: Realtek Semiconductor Co., Ltd. RTL8169 PCI Gigabit Ethernet Controller [10ec:8169] (rev 10) Subsystem: Realtek Semiconductor Co., Ltd. RTL8169/8110 Family PCI Gigabit Ethernet NIC [10ec:8169] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV+ VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- SERR- Kernel driver in use: r8169 Kernel modules: r8169 02:0a.0 USB controller [0c03]: VIA Technologies, Inc. VT82xx/62xx UHCI USB 1.1 Controller [1106:3038] (rev 62) (prog-if 00 [UHCI]) Subsystem: VIA Technologies, Inc. VT82xx/62xx UHCI USB 1.1 Controller [1106:3038] Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR+ FastB2B- DisINTx- Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=medium >TAbort- SERR- Kernel driver in use: uhci_hcd Kernel modules: uhci_hcd 02:0a.1 USB controller [0c03]: VIA Technologies, Inc.
