pie stuff

pie stuff

[Elena Zannoni]

A heads up, I am going to start committing pie support to my old
branch (which I am bringing up to date) in a few hours.  I will then
merge it piecemeal to mainline. It's not that much code. More details
later.

elena

Re: pie stuff

[Stan Shebs]

Elena Zannoni wrote:

>A heads up, I am going to start committing pie support to my old
>branch (which I am bringing up to date) in a few hours.  I will then
>merge it piecemeal to mainline. It's not that much code. More details
>later.
>
I must not have gotten the memo - what's pie? (yeah, yeah, filled pastry
that one eats :-) )

Stan

Re: pie stuff

[Elena Zannoni]

Stan Shebs writes:
 > Elena Zannoni wrote:
 > 
 > >A heads up, I am going to start committing pie support to my old
 > >branch (which I am bringing up to date) in a few hours.  I will then
 > >merge it piecemeal to mainline. It's not that much code. More details
 > >later.
 > >
 > I must not have gotten the memo - what's pie? (yeah, yeah, filled pastry
 > that one eats :-) )
 > 
 > Stan

exactly, next will be gdb brewing your own tea. Tea, earl grey,
hot. An appetizing combinations. And who says we never think about the
comfort of the poor users?

It's position independent executables. Your main exec is just like a
shared library. Every time you run it you get a different start
address. Not fun, I can tell you. gcc -fpie -pie -g -o blah blah.c

elena

Re: pie stuff

[Christopher Faylor]

On Tue, Mar 23, 2004 at 02:20:10PM -0800, Stan Shebs wrote:
>Elena Zannoni wrote:
>>A heads up, I am going to start committing pie support to my old branch
>>(which I am bringing up to date) in a few hours.  I will then merge it
>>piecemeal to mainline.  It's not that much code.  More details later.
>>
>I must not have gotten the memo - what's pie?  (yeah, yeah, filled
>pastry that one eats :-) )

Position Independent Executable

cgf

Re: pie stuff

[Stan Shebs]

Elena Zannoni wrote:

>Stan Shebs writes:
> > Elena Zannoni wrote:
> > 
> > >A heads up, I am going to start committing pie support to my old
> > >branch (which I am bringing up to date) in a few hours.  I will then
> > >merge it piecemeal to mainline. It's not that much code. More details
> > >later.
> > >
> > I must not have gotten the memo - what's pie? (yeah, yeah, filled pastry
> > that one eats :-) )
> > 
> > Stan
>
>exactly, next will be gdb brewing your own tea. Tea, earl grey,
>hot. An appetizing combinations. And who says we never think about the
>comfort of the poor users?
>
>
And not even implausible! GDB could inject itself into the embedded
systems that are in modern coffeemakers, through the stub into
flash memory. Why run an *inferior* process when you can replace
it with a *superior* debugger?

>It's position independent executables. Your main exec is just like a
>shared library. Every time you run it you get a different start
>address. Not fun, I can tell you. gcc -fpie -pie -g -o blah blah.c
>
Heh heh, reminiscent of AIX hell...

Stan

Re: pie stuff

[Eli Zaretskii]

> From: Elena Zannoni <ezannoni@redhat.com>
> Date: Tue, 23 Mar 2004 17:33:16 -0500
> 
> It's position independent executables. Your main exec is just like a
> shared library. Every time you run it you get a different start
> address. Not fun, I can tell you. gcc -fpie -pie -g -o blah blah.c

Whoever came up with that monstrous idea, I wonder... ;-)

Re: pie stuff

[Andrew Cagney]

>>> From: Elena Zannoni <ezannoni@redhat.com>
>>> Date: Tue, 23 Mar 2004 17:33:16 -0500
>>> 
>>> It's position independent executables. Your main exec is just like a
>>> shared library. Every time you run it you get a different start
>>> address. Not fun, I can tell you. gcc -fpie -pie -g -o blah blah.c
> 
> 
> Whoever came up with that monstrous idea, I wonder... ;-)

(I'm not going there)

Anyway, I think this is something for PROBLEMS - broken in 6.1, fixes 
available for mainline.

Andrew

Re: pie stuff

[Eli Zaretskii]

> Date: Wed, 24 Mar 2004 09:44:35 -0500
> From: Andrew Cagney <cagney@gnu.org>
> > 
> > Whoever came up with that monstrous idea, I wonder... ;-)
> 
> (I'm not going there)
> 
> Anyway, I think this is something for PROBLEMS - broken in 6.1, fixes 
> available for mainline.

Agreed.