On Mon, Aug 27, 2018 at 6:34 AM H. Peter Anvin wrote:
>
> On 08/27/18 06:11, Peter Zijlstra wrote:
> > On Mon, Aug 27, 2018 at 05:26:53AM -0700, H. Peter Anvin wrote:
> >
> >> _THIS_IP_, however, is completely ill-defined, other than being an
> >> address *somewhere* in the same global function (n
On 08/27/18 06:11, Peter Zijlstra wrote:
> On Mon, Aug 27, 2018 at 05:26:53AM -0700, H. Peter Anvin wrote:
>
>> _THIS_IP_, however, is completely ill-defined, other than being an
>> address *somewhere* in the same global function (not even necessarily
>> the same function if the function is static
On Mon, Aug 27, 2018 at 05:26:53AM -0700, H. Peter Anvin wrote:
> _THIS_IP_, however, is completely ill-defined, other than being an
> address *somewhere* in the same global function (not even necessarily
> the same function if the function is static!) As my experiment show, in
> many (nearly) ca
On 08/27/18 00:33, Peter Zijlstra wrote:
>
> What problem are we trying to solve? _THIS_IP_ and _RET_IP_ work fine.
> We're 'good' at dealing with text addresses, we use them for call stacks
> and all sorts. Why does this need changing?
>
_RET_IP_ works fine, with the following two caveats:
1.
[ Trimmed the cc list because my SMTP didn't accept that many
addresses. ]
On Sun, 26 Aug 2018 13:25:14 -0700
Linus Torvalds wrote:
> On Sun, Aug 26, 2018 at 12:32 PM H. Peter Anvin wrote:
> >
> > Here is a full-blown (user space) test program demonstrating the whole
> > technique and how to us
On Sun, Aug 26, 2018 at 07:52:59PM -0700, Nick Desaulniers wrote:
> On Sun, Aug 26, 2018 at 1:25 PM Linus Torvalds
> > Instead, maybe we could encourage something like
> >
> > struct kernel_loc { const char *file; const char *fn; int line; };
> >
> > #define __GEN_LOC__(n) \
> > ({ sta
On Sun, Aug 26, 2018 at 1:25 PM Linus Torvalds
wrote:
> Honestly, I'd suggest:
>
> - just do the current_text_addr() to _THIS_IP_ conversion
>
> - keep _THIS_IP_ and make it be the generic one, and screw the whole
> "some architectures might implement is better" issue. Nobody cares.
And mention
On 08/26/18 12:30, H. Peter Anvin wrote:
> Here is a full-blown (user space) test program demonstrating the whole
> technique and how to use it.
>
> -hpa
Incidentally, it looks like _RET_IP_ really should be defined as:
/*
* Is there any reason whatsoever to have _RET_IP_ an unsigned int
On Sun, Aug 26, 2018 at 12:32 PM H. Peter Anvin wrote:
>
> Here is a full-blown (user space) test program demonstrating the whole
> technique and how to use it.
So while I agree that some _THIS_IP_ users might be better off being
converted to __builtin_return_address(0) at the caller, I also thin
Here is a full-blown (user space) test program demonstrating the whole
technique and how to use it.
-hpa
#include
#include
#define _RET_IP_ ((unsigned long)__builtin_return_address(0))
#define noinline __attribute__((noinline))
#define used __attribute__((used))
/* __always_inline is d
On 08/25/18 20:16, H. Peter Anvin wrote:
> On 08/25/18 19:38, H. Peter Anvin wrote:
>>
>> If it was worthwhile it would make more sense to at least force this
>> into the rodata section with the string, something like the attached
>> file for an example; however, I have a hunch it doesn't matter.
>
On 08/25/18 19:38, H. Peter Anvin wrote:
>
> If it was worthwhile it would make more sense to at least force this
> into the rodata section with the string, something like the attached
> file for an example; however, I have a hunch it doesn't matter.
>
An even nuttier version which avoids the ext
On 08/25/18 03:48, Helge Deller wrote:
>
> Currently alpha, s390, sparc, sh, c6x, ia64 and parisc provide an
> inline assembly function to get the current instruction pointer.
> As mentioned in an earlier thread, I personally would *prefer* if
> _THIS_IP_ would use those inline assembly instruct
On Tue, Aug 21, 2018 at 1:31 PM Nick Desaulniers
wrote:
>
> I suspect that current_text_addr predated GNU C extensions for statement
> expressions and/or taking the address of a label, then the macro was
> reimplemented for every new archs include/asm/processor.h, even though
> there were very few
On 21.08.2018 22:28, Nick Desaulniers wrote:
> Prefer _THIS_IP_ defined in linux/kernel.h.
>
> Most definitions of current_text_addr were the same as _THIS_IP_, but
> a few archs had inline assembly instead.
>
> This patch removes the final call site of current_text_addr, making all
> of the defi
15 matches
Mail list logo
