Fedora Account System
Red Hat Associate
Red Hat Customer
Created attachment 2122381 [details] gzip's flat file containing the log of system activity up to the freeze/crash. 1. Please describe the problem: Running a locally built but otherwise NOT MODIFIED Linux kernel "linux-6.19-rc5" and building the "latest" gcc from the gcc repo generates msg "kernel BUG at kernel/irq_work.c:245!". The rc4 version of this kernel does NOT fail. The build process is the SAME in both cases. 2. What is the Version-Release number of the kernel: linux-6.19-rc5 3. Did it work previously in Fedora? If so, what kernel version did the issue *first* appear? Old kernels are available for download at https://koji.fedoraproject.org/koji/packageinfo?packageID=8 : This problem appeared in the rc5 version of the kernel, the rc4 version of this kernel does NOT fail 4. Can you reproduce this issue? If so, please provide the steps to reproduce the issue below: Yes Build gcc and gmp, mpfr, isl, and mpc. The build log up to the point of the freeze/crash is available for this situation. 5. Does this problem occur with the latest Rawhide kernel? To install the Rawhide kernel, run ``sudo dnf install fedora-repos-rawhide`` followed by ``sudo dnf update --enablerepo=rawhide kernel``: Not as far as I can tell. 6. Are you running any modules that not shipped with directly Fedora's kernel?: No 7. Please attach the kernel logs. You can get the complete kernel log for a boot with ``journalctl --no-hostname -k > dmesg.txt``. If the issue occurred on a previous boot, use the journalctl ``-b`` flag.
Howdy, Yes. I know. Fedora (Rawhide) kernel 6.19.0-rc[4-5] are not yet in the Fedora repository. We will be soon and I thought I'd make a report just in case anyone is interested. Would it be possible to get this info transferred upstream by any chance? I'm having some trouble figuring out which mailing list to send my report to. Best regards, George...
Hey George, Are you able to reproduce this issue with the most recent Fedora 44 as well as the latest nightly build? https://openqa.fedoraproject.org/nightlies.html If so, are you able to provide fresh sets of the kernel logs? While it is not necessary right now, would you be amenable to setting up kdump as well? I ask because you opened an extremely similar bug at https://bugzilla.redhat.com/show_bug.cgi?id=2432311 and I have a hunch it is the same issue just manifesting in different ways. If the system is indeed locking up such that the kernel is not even logging anything, then a vmcore (dump of kernel memory) may be necessary to see what's going on. For this, you will additionally need to setup sysrq to allow for crashing while the system is locked up. You can set kernel.sysrq=1 and then Alt + PrintScn + t and check dmesg if a thread dump occurred to check it worked. If the system is properly locking up, however, where interrupts aren't even being handled, then even this may not work.