public inbox for gcc-help@gcc.gnu.org
 help / color / mirror / Atom feed
* Optimization auto-setting _FORTIFY_SOURCE ??
@ 2023-01-29 23:05 Patrick Herbst
  2023-01-30  4:23 ` Xi Ruoyao
  0 siblings, 1 reply; 3+ messages in thread
From: Patrick Herbst @ 2023-01-29 23:05 UTC (permalink / raw)
  To: gcc-help

Using GCC 9.4, I've noticed setting -O1 will set
_FORTIFY_SOURCE=2

I'm able to see this by running
gcc -c -E -dM -O1 - < /dev/null | grep FORTIFY_SOURCE

But running the same command without -O1 doesn't set it.

Is this documented?  is this expected behavior?  I could not find any
reference to how optimization leads to setting this macro.

Any tips on where I can look, or what i'm missing to see this connection?

Thanks!

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: Optimization auto-setting _FORTIFY_SOURCE ??
  2023-01-29 23:05 Optimization auto-setting _FORTIFY_SOURCE ?? Patrick Herbst
@ 2023-01-30  4:23 ` Xi Ruoyao
  2023-01-30  4:57   ` LIU Hao
  0 siblings, 1 reply; 3+ messages in thread
From: Xi Ruoyao @ 2023-01-30  4:23 UTC (permalink / raw)
  To: Patrick Herbst; +Cc: gcc-help

On Sun, 2023-01-29 at 18:05 -0500, Patrick Herbst via Gcc-help wrote:
> Using GCC 9.4, I've noticed setting -O1 will set
> _FORTIFY_SOURCE=2
> 
> I'm able to see this by running
> gcc -c -E -dM -O1 - < /dev/null | grep FORTIFY_SOURCE
> 
> But running the same command without -O1 doesn't set it.
> 
> Is this documented?  is this expected behavior?  I could not find any
> reference to how optimization leads to setting this macro.
> 
> Any tips on where I can look, or what i'm missing to see this
> connection?

It's because your distro hacked GCC code to define _FORTIFY_SOURCE when
optimization is enabled.  Vanilla GCC does not have this behavior.


-- 
Xi Ruoyao <xry111@xry111.site>
School of Aerospace Science and Technology, Xidian University

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: Optimization auto-setting _FORTIFY_SOURCE ??
  2023-01-30  4:23 ` Xi Ruoyao
@ 2023-01-30  4:57   ` LIU Hao
  0 siblings, 0 replies; 3+ messages in thread
From: LIU Hao @ 2023-01-30  4:57 UTC (permalink / raw)
  To: Xi Ruoyao, Patrick Herbst; +Cc: gcc-help


[-- Attachment #1.1: Type: text/plain, Size: 1230 bytes --]

在 2023/1/30 12:23, Xi Ruoyao via Gcc-help 写道:
> On Sun, 2023-01-29 at 18:05 -0500, Patrick Herbst via Gcc-help wrote:
>> Using GCC 9.4, I've noticed setting -O1 will set
>> _FORTIFY_SOURCE=2
>>
>> I'm able to see this by running
>> gcc -c -E -dM -O1 - < /dev/null | grep FORTIFY_SOURCE
>>
>> But running the same command without -O1 doesn't set it.
>>
>> Is this documented?  is this expected behavior?  I could not find any
>> reference to how optimization leads to setting this macro.
>>
>> Any tips on where I can look, or what i'm missing to see this
>> connection?
> 
> It's because your distro hacked GCC code to define _FORTIFY_SOURCE when
> optimization is enabled.  Vanilla GCC does not have this behavior.
> 
> 

For the record: Debian and Ubuntu are known distros that have `_FORTIFY_SOURCE` set when 
optimization is enabled. The relevant patch is here

    https://sources.debian.org/src/gcc-8/8.3.0-2/debian/patches/gcc-default-fortify-source.diff/

however the hunk against 'invoke.texi' is incorrect: It says `_FORTIFY_SOURCE` is enabled if 
optimization level 'is set to 2 or higher', but it is actually defined for any optimization level > 0.

-- 
Best regards,
LIU Hao


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 840 bytes --]

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2023-01-30  4:57 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-01-29 23:05 Optimization auto-setting _FORTIFY_SOURCE ?? Patrick Herbst
2023-01-30  4:23 ` Xi Ruoyao
2023-01-30  4:57   ` LIU Hao

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).