On Wed, Jan 03, 2024 at 09:43:32AM +0100, Mathieu Othacehe wrote: > > Hello Efraim, > > > + ;; On some architectures 'struct sigcontext' gets redefined from > > + ;; %linux-libre-headers/include/asm/sigcontext.h > > + ,@(cond ((target-mips64el?) > > + `(#:phases (modify-phases %standard-phases > > + (add-before 'configure 'patch-mips-old-h > > + (lambda _ > > + (substitute* "src/fault-linux-mips-old.h" > > + (("#include ") ""))))))) > > + ((target-x32?) > > + `(#:phases (modify-phases %standard-phases > > + (add-before 'configure 'patch-linux-x86_64-old-h > > + (lambda _ > > + (substitute* "src/fault-linux-x86_64-old.h" > > + (("#include ") ""))))))) > > + (else '())))) > > (description > > "GNU libsigsegv is a library to handle page faults, which occur when a > > program tries to access an unavailable region of memory, in user mode. By > > #67859 is doing more or less the same thing, but for x86_64 as a > whole. Could you maybe replace `target-x32` by `target-x86`? > > Maybe we should come up with a list of architectures that need that kind > of fix in libsigsegv and patch all the src/fault-linux-xxx-old-h? I don't have a mips machine hooked up so I can't test if its needed natively. I did check debian's packaging of libsigsegv and they don't have anything like this for any architecture. It wouldn't be hard to change it to (if (%current-target-system) ... (substitute* (find-files "src" "fault.*-old\\.h$") which would hit every instance of this include > The rest of the series seems fine to me. > > Thanks, > > Mathieu -- Efraim Flashner רנשלפ םירפא GPG key = A28B F40C 3E55 1372 662D 14F7 41AA E7DC CA3D 8351 Confidentiality cannot be guaranteed on emails sent or received unencrypted