On Tue, 2022-09-20 at 05:44 +, Christophe Leroy wrote:
>
> As far as I know, cachelines are minimum 64 bytes on PPC64 aren't
> they ?
In practice maybe. I don't know what the convention is in the kernel in
cases where the ISA is less specific than what the supported machines
do.
> > Related
Le 20/09/2022 à 04:32, Benjamin Gray a écrit :
> On Mon, 2022-09-19 at 07:16 +, Christophe Leroy wrote:
>> Why would it be unpredictable ? Only one page is mapped. If it
>> crosses
>> the boundary, __put_kernel_nofault() will fail in a controled manner.
>> I see no point in doing the check be
On Mon, 2022-09-19 at 07:16 +, Christophe Leroy wrote:
> Why would it be unpredictable ? Only one page is mapped. If it
> crosses
> the boundary, __put_kernel_nofault() will fail in a controled manner.
> I see no point in doing the check before every write.
Oh I didn't see that get_vm_area au
On Mon, 2022-09-19 at 06:38 +, 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))
> > +
Le 19/09/2022 à 08:49, Benjamin Gray a écrit :
> On Mon, 2022-09-19 at 06:04 +, Christophe Leroy wrote:
>> With CONFIG_STRICT_KERNEL_RWX, this patches causes a 15% time
>> increase
>> for activation/deactivation of ftrace.
>
> It's possible that new alignment check is the cause. I'll see
>
On Mon, 2022-09-19 at 06:04 +, Christophe Leroy wrote:
> With CONFIG_STRICT_KERNEL_RWX, this patches causes a 15% time
> increase
> for activation/deactivation of ftrace.
It's possible that new alignment check is the cause. I'll see
> Without CONFIG_STRICT_KERNEL_RWX, it doesn't build.
Yup
Le 16/09/2022 à 08:23, Benjamin Gray a écrit :
>
> -static int do_patch_instruction(u32 *addr, ppc_inst_t instr)
> +static int patch_text(void *dest, const void *src, size_t size, bool is_exec)
> {
> int err;
> unsigned long flags;
>
> - /*
> - * During early early bo
Le 16/09/2022 à 08:23, Benjamin Gray a écrit :
> Adds a generic text patching mechanism for patches of 1, 2, 4, or 8
> bytes. The patcher conditionally syncs the icache depending on if
> the content will be executed (as opposed to, e.g., read-only data).
>
> The `patch_instruction` function is r
Hi Benjamin,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on powerpc/next]
[also build test WARNING on next-20220916]
[cannot apply to linus/master v6.0-rc5]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we su
Hi Benjamin,
Thank you for the patch! Yet something to improve:
[auto build test ERROR on powerpc/next]
[also build test ERROR on next-20220916]
[cannot apply to linus/master v6.0-rc5]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to
Adds a generic text patching mechanism for patches of 1, 2, 4, or 8
bytes. The patcher conditionally syncs the icache depending on if
the content will be executed (as opposed to, e.g., read-only data).
The `patch_instruction` function is reimplemented in terms of this
more generic function. This g
11 matches
Mail list logo
