On Wed, Jan 18, 2023 at 5:03 AM Petr Mladek <pmladek@xxxxxxxx> wrote: > [...] > > case R_X86_64_PC32: > > case R_X86_64_PLT32: > > - if (*(u32 *)loc != 0) > > - goto invalid_relocation; > > - val -= (u64)loc; > > - write(loc, &val, 4); > > #if 0 > > - if ((s64)val != *(s32 *)loc) > > + if ((s64)val != *(s32 *)&val) > > goto overflow; > > This is supposed to check the to-be-written value. Great catch! Fixed it in v9. > > > #endif > > + val -= (u64)loc; > > This is modifying the to-be-written value. It should be computed before > the overflow check. > > I know that the check is not really compiled in but we should > not break it. > > > > break; > > Otherwise, it looks fine. > > > Now, I agree with Miroslav that we should get an approval from x86 > maintainers. Sigh, I think that I have already asked for this earlier: > > !!! Please add x86@xxxxxxxxxx and linux-kernel@xxxxxxxxxxxxxxx at > minimum into CC when sending V9 !!! I am sorry I missed this request. > > The more people know about this change the better. And it is really > important to make maintainers of the touched subsystem aware of > proposed changes. > > It is a good practice to add people that are printed by > ./scripts/get_maintainer.pl. In this case, it is: > > $> ./scripts/get_maintainer.pl arch/x86/kernel/module.c > Thomas Gleixner <tglx@xxxxxxxxxxxxx> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT),authored:2/12=17%,added_lines:24/74=32%,removed_lines:5/26=19%) > Ingo Molnar <mingo@xxxxxxxxxx> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) > Borislav Petkov <bp@xxxxxxxxx> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT),commit_signer:4/12=33%) > Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx> (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) > x86@xxxxxxxxxx (maintainer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) > "H. Peter Anvin" <hpa@xxxxxxxxx> (reviewer:X86 ARCHITECTURE (32-BIT AND 64-BIT)) > Peter Zijlstra <peterz@xxxxxxxxxxxxx> (commit_signer:8/12=67%,authored:4/12=33%,added_lines:41/74=55%,removed_lines:8/26=31%) > Kees Cook <keescook@xxxxxxxxxxxx> (commit_signer:4/12=33%) > Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> (commit_signer:3/12=25%) > "Jason A. Donenfeld" <Jason@xxxxxxxxx> (commit_signer:3/12=25%,authored:3/12=25%,removed_lines:3/26=12%) > Julian Pidancet <julian.pidancet@xxxxxxxxxx> (authored:1/12=8%,added_lines:5/74=7%,removed_lines:6/26=23%) > Ard Biesheuvel <ardb@xxxxxxxxxx> (authored:1/12=8%,removed_lines:3/26=12%) > linux-kernel@xxxxxxxxxxxxxxx (open list:X86 ARCHITECTURE (32-BIT AND 64-BIT)) vger.kernel.org usually drops my email when the CC list is so long. I am trying to fix it. For v9, I will CC x86@xxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, and linux-modules@xxxxxxxxxxxxxxx. Thanks, Song
