This bug is awaiting verification that the linux/5.4.0-91.102 kernel in -proposed solves the problem. Please test the kernel and update this bug with the results. If the problem is solved, change the tag 'verification-needed-focal' to 'verification-done-focal'. If the problem still exists, change the tag 'verification-needed-focal' to 'verification-failed-focal'.
If verification is not done by 5 working days from today, this fix will be dropped from the source code, and this bug will be closed. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Thank you! ** Tags added: verification-needed-bionic -- You received this bug notification because you are a member of Kernel Packages, which is subscribed to linux in Ubuntu. https://bugs.launchpad.net/bugs/1948351 Title: ppc64 BPF JIT mod by 1 will not return 0 Status in linux package in Ubuntu: Incomplete Status in linux source package in Bionic: Fix Committed Status in linux source package in Focal: Fix Committed Status in linux source package in Hirsute: Fix Committed Status in linux source package in Impish: Fix Committed Bug description: [Impact] When doing MOD by 1 with a immediate/constant divisor on PPC, the JIT will produce code that returns the dividend, just like a division, instead of 0. Both eBPF and cBPF will fail as well when doing such operations. [Test case] $ cat bpf-mod1.c #include <linux/bpf.h> #include <unistd.h> #include <sys/syscall.h> #include <sys/socket.h> #include <stdio.h> #include <sys/prctl.h> #include <stdlib.h> #include <errno.h> #include <limits.h> #include <err.h> #include <string.h> #include <linux/filter.h> #define ARRAY_SIZE(array) (sizeof(array)/sizeof(array[0])) static int pair[2]; static int attach() { int r; struct sock_filter insn[] = { { BPF_LD | BPF_W | BPF_ABS, 0, 0, 0 }, { BPF_ALU | BPF_MOD, 0, 0, 1 }, { BPF_RET | BPF_A, 0, 0, 0 }, }; struct sock_fprog prog = {}; prog.filter = insn; prog.len = ARRAY_SIZE(insn); socketpair(AF_UNIX, SOCK_DGRAM, 0, pair); setsockopt(pair[1], SOL_SOCKET, SO_ATTACH_FILTER, &prog, sizeof(prog)); return 0; } int main(int argc, char **argv) { int buf[5]; int r; r = attach(); if (r < 0) { err(1, "function will error out already"); } write(pair[0], "hello", 5); r = recv(pair[1], buf, 5, MSG_DONTWAIT); if (r != -1 || errno != EAGAIN) { err(1, "program failed"); } return 0; } $ gcc -o bpf-mod1 bpf-mod1.c $ ./bpf-mod1 cbpf-mod1: program failed: Success After fix: $ ./bpf-mod1 $ echo $? 0 [Potential regression] BPF programs might be misbehave on ppc64el. ======================================================== This is a scripted bug report about ADT failures while running linux tests for linux/5.4.0-90.101 on focal. Whether this is caused by the dep8 tests of the tested source or the kernel has yet to be determined. Consistently failing on Focal/linux 5.4.0-90.101 13:15:26 DEBUG| [stdout] # selftests: net: test_bpf.sh 13:15:27 DEBUG| [stdout] # test_bpf: [FAIL] 13:15:27 DEBUG| [stdout] not ok 9 selftests: net: test_bpf.sh # exit=1 Testing failed on: ppc64el: https://autopkgtest.ubuntu.com/results/autopkgtest-focal/focal/ppc64el/l/linux/20211021_141544_eda49@/log.gz To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1948351/+subscriptions -- Mailing list: https://launchpad.net/~kernel-packages Post to : kernel-packages@lists.launchpad.net Unsubscribe : https://launchpad.net/~kernel-packages More help : https://help.launchpad.net/ListHelp
