On 07/28/2017 01:21 PM, Richard Biener wrote:
> On Fri, Jul 28, 2017 at 12:52 PM, Martin Liška wrote:
>> On 07/28/2017 09:58 AM, Richard Biener wrote:
>>> On Fri, Jul 28, 2017 at 9:50 AM, Martin Liška wrote:
On 07/28/2017 09:21 AM, Richard Biener wrote:
> On Thu, Jul 27, 2017 at 4:24 PM,
This patch adds aarch64 supports for stack clash mitigation.
Changes since V2:
Uses new PARAMS to control the guard size and probing interval. Uses a
64k guard by default and a 1k outgoing args size.
Simplified code slightly as a result of "can't happen" cases.
Probe the last allocated word i
This contains the PPC bits for stack clash protection.
Changes since V2:
Exploits inlined/unrolled probes and rotated loops for the dynamic area.
Some trivial simplifications. It also uses the new params to control
if probes are needed and how often to probe.
Jeff
* config/rs6000/rs6
These are the s390 patches for stack clash mitigation.
Changes since V2:
Incorporate changes from IBM to address missing functionality.
Hopefully no longer regresses Ada as we no longer define
STACK_CHECK_STATIC_BUILTIN.
Jeff
* config/s390/s390.c (MIN_UNROLL_PROBES): Define.
(
This patch catches a couple places where the generic parts of the
compiler would transform prologue probing sequences in unsafe ways.
Changes since V2:
Testing for aarch64 showed the scheduler would perform transformations
similar to combine-stack-adjustments via its dependency breaking
transfor
This patch adds stack clash protection to the x86 port and provides the
first set of tests that will be used by the other ports in the future.
Changes since V2:
This uses the new PARAMs to determine if probing is needed and how often
to probe once probing is needed. New tests were added to veri
This patch introduces some routines for logging of stack clash
protection actions.
I don't think this patch changed at all relative to V2.
Jeff
* function.c (dump_stack_clash_frame_info): New function.
* function.h (dump_stack_clash_frame_info): Prototype.
(enum stack_cl
This patch adds checking of flag_stack_protection to various backends
that will rely on partial protection against stack-clash via -fstack-check.
I don't think this patch changed at all relative to V2.
Jeff
* config/alpha/alpha.c (alpha_expand_prologue): Also check
flag_stack_clas
This patch introduces generic mechanisms to protect the dynamically
allocated stack space against stack-clash attacks.
Changes since V2:
Dynamic allocations can be emitted as unrolled inlined probes or with a
rotated loop. Blockage insns are also properly emitted for the dynamic
area probes and
This patch introduces the stack clash protection options
Changes since V2:
Adds two new params. The first controls the size of the guard area.
This controls the threshold for when a function prologue requires
probes. The second controls the probing interval -- ie, once probes are
needed, how of
OK, so about a week later than I wanted. Too many fires, not enough
water. The V3 patch has expanded a bit...
1. For constant sized dynamic allocations we'll allocate/probe up to 4
STACK_CLASH_PROTECTION_PROBE_INTERVAL regions inline and unrolled.
2. For larger constant sized dynam
On 07/30/2017 11:55 AM, Thomas Koenig wrote:
> Hello world,
>
> the attached patch performs a runtime-check for the dim arguments
> for transformational intrinsic functions.
>
> I debated with myself a bit if I should make this depend on
> bounds checking, but I finally decided against it. The t
Hello!
attribute ((naked)) generates function body without function frame,
and as shown in PR 25967 [1], users are looking for this feature also
for x86 targets. Recently, Daniel introduced a testcase that would
benefit from this attribute.
Attached patch also adds a trap instruction (ud2) at the
Hi,
PR81354 identifies a latent bug that can happen in SLSR since the
conditional candidate support was first added. SLSR relies on the
address of a GIMPLE PHI remaining constant during the course of the
optimization pass, but it needs to split edges. The use of
make_single_succ_edge and reinsta
> On Jul 28, 2017, at 4:08 PM, Segher Boessenkool
> wrote:
>
> Hi!
>
> On Thu, Jul 27, 2017 at 07:21:14PM -0400, Michael Meissner wrote:
>> This patches optimizes the PowerPC vector set operation for 64-bit doubles
>> and
>> longs where the elements in the vector set may have been extracted f
2017-07-28 22:40 GMT+02:00 Daniel Krügler :
> 2017-07-28 22:29 GMT+02:00 Daniel Krügler :
>> 2017-07-28 22:25 GMT+02:00 Tim Song :
>>> On Fri, Jul 28, 2017 at 4:10 PM, Daniel Krügler
>>> wrote:
+ // Performs an implicit conversion from _Tp to __sv_type.
+ template
+
On Wed, Jul 26, 2017 at 2:18 PM, Uros Bizjak wrote:
> Hello!
>
> This patch is the prerequisite for my next patch. It enables %z
> extension for asm_fprintf that will emit 'q' or 'l' suffixes for
> instructions with word-sized operands.
>
> 2017-07-26 Uros Bizjak
>
> * c-format.c (asm_fprin
On Sat, Jul 29, 2017 at 4:22 PM, H.J. Lu wrote:
> Since there is an extra error code passed to the exception handler,
> INCOMING_FRAME_SP_OFFSET is return address plus error code for the
> exception handler. This patch updates INCOMING_FRAME_SP_OFFSET to
> the correct value for the exception hand
On Sat, Jul 29, 2017 at 3:06 PM, H.J. Lu wrote:
> Add a new file, avx512-check.h, to check all AVX512 features. The test
> is skipped if any requested AVX512 features are unavailable.
>
> Tested on Skylake server and Haswell. OK for trunk?
No, I'd rather leave it in in the way they are now, so
Hi all,
This is an updated version of patch in
https://gcc.gnu.org/ml/gcc-patches/2017-07/msg00924.html . It prevents
optimization in presense of sNaNs (and qNaNs when comparison operator is
> >= < <=) to preserve FP exceptions.
I've fixed some stylistic issues mentioned in previous review.
20 matches
Mail list logo
