On Mon, 2022-09-19 at 06:38 +0000, Christophe Leroy wrote: > Le 16/09/2022 à 08:23, Benjamin Gray a écrit : > > [...] > > + /* Make sure we aren't patching a freed init section */ > > + if (static_branch_likely(&init_mem_is_free) && > > init_section_contains(dest, 4)) > > + return 0; > > + > > if (!static_branch_likely(&poking_init_done)) > > - return raw_patch_instruction(addr, instr); > > + return __patch_text(dest, src, size, is_exec, > > dest); > > > > Test ordering looks odd. How can init_mem_is_free be true and > poking_init_done be false ? > > That's just the flattened result of the original STRICT_KERNEL_RWX case (as I hadn't tested building with it disabled).
With the build fixed for v2, the structure is similar to the original the second check is not performed if CONFIG_STRICT_KERNEL_RWX is disabled. >
