Hi,
I did indeed create a trace file, and it's huge. The check is correct, but
the values are false. The memblockbase is fetched from
memblock.memory.regions[i].base, which has been initialized with a wrong
value (0x1), which is the same as top.
I have not yet tracked down where memblock.memo
Yes, thats correct. The amount of free memory on the system is around 4G
(there is a total of 6G on the system).
On Fri, Dec 16, 2011 at 12:26 AM, Mahmood Naderan wrote:
> Do you mean the "free" column in "free -m" shows 4G?
>
> The amount of memory you specify in AddrRange("512MB") should always
Hi Tony, thank you. I download the most new copy of gem5. When I run with the
command ./build/ARM_FS/gem5.opt ./configs/example/fs.py -b ArmAndroid, it
cannot open the disk file. The benchmark BBench comes into the same problem.
gem5 Simulator System. http://gem5.org
gem5 is copyrighted softwar
Do you mean the "free" column in "free -m" shows 4G?
The amount of memory you specify in AddrRange("512MB") should always
be less than *free* available memory in your host. For example, if you
set AddrRange to 512MB and have 4G RAM installed on your host but have
only 200MB free memory, then you g
Also, if you haven't seen it, this page documents some of the methods
you can use to debug what's going on.
http://www.gem5.org/Debugging
The section called "Trace-based debugging" will probably be the most
useful. Be sure to only trace a portion of execution at a time. Tracing
too much creates s
Where bottom >= top, is bottom really greater than or equal to top? Or
in other words, is the comparison returning the wrong result, or are the
values wrong in the first place? If it's the comparison that's wrong,
that would be a bit surprising but easier to debug because it's pretty
contained. If
Someone should just wrap the declarations with this in the header
(don't wrap the #include statement though. I'd do it, but I'm mobile.
#ifdef __cplusplus
extern "C" {
#endif
#ifdef __cplusplus
}
#endif
On Thu, Dec 15, 2011 at 8:43 AM, Ali Saidi wrote:
> extern "C"?
>
> Ali
>
>
> On Thu, 15 De
Yes, it should work. Thanks, Ali.
On Thu, 15 Dec 2011 10:43:47 -0600, Ali Saidi wrote:
> extern "C"?
>
> Ali
>
> On Thu, 15 Dec 2011 10:20:39 -0500, wrote:
>> When I tried to use m5_dump_stats and m5_reset_stats in my program,
>> the
>> functions are written in assembly which is only reco
That's cool. I though maybe extending the PCEvent class with a PyPCEvent
class which should have a callback to python. But before I can do that I'll
have to read up on the swig stuff, which I am not that familiar with.
On Thu, Dec 15, 2011 at 5:43 PM, Ali Saidi wrote:
> You can add events like t
You need to add bbench to your ./configs/common/Benchmarks.py file. If you
get the latest changeset this is already done.
-Tony
On Thu, Dec 15, 2011 at 11:09 AM, yanke wrote:
> Hi Tony, I'am sorry so late to reply you. I just follow the step at
> http://www.gem5.org/Bbench-gem5. When I run ./b
extern "C"?
Ali
On Thu, 15 Dec 2011 10:20:39 -0500, wrote:
When I tried to use m5_dump_stats and m5_reset_stats in my program,
the
functions are written in assembly which is only recognized by gcc,
not by
g++. So I cannot use these funcs when my source code is C++. I just
wonder
if someone c
You can add events like the following:
if
(kernelSymtab->findAddress("__udelay", addr)) {
uDelaySkipEvent =
new UDelayEvent(&pcEventQueue, "__udelay", fixFuncEventAddr(addr), 1000,
0);
}
This could be based on parameters and will work just fine if
you are interested in kernel events.
free -m indicates that I have around 4G of memory available on my system.
On Thu, Dec 15, 2011 at 3:40 AM, Mahmood Naderan wrote:
> how much free memory do you have on host?
> "free -m" will show that
>
> On 12/15/11, Ankita (Garg) Goel wrote:
> > I increased it to 4096MB, but still no luck. Thi
Hi Tony, I'am sorry so late to reply you. I just follow the step at
http://www.gem5.org/Bbench-gem5. When I run ./build/ARM_FS/gem5.opt
./configs/examples/fs.py -b ArmAndroid, I can't boot into vnc. When I run
./build/ARM_FS/gem5.opt ./configs/examples/fs.py -b bbench, it says there is no
bbenc
When I tried to use m5_dump_stats and m5_reset_stats in my program, the
functions are written in assembly which is only recognized by gcc, not by
g++. So I cannot use these funcs when my source code is C++. I just wonder
if someone can help give me any hint or even a link to where I can rewrite
th
Hi,
To my understanding, there are no interfaces to monitor virtual PC
addresses provided by gem5. The most convenient way is instrumentation.
-Yongbing Huang
From: gem5-users-boun...@gem5.org [mailto:gem5-users-boun...@gem5.org] On
Behalf Of Anders Handler
Sent: Thursday, Dece
Hi,
I am making full system simulations on X86. What I would like to do is to
make "breakpoints" which can start and stop statistics and dump trace to a
file.
My breakpoints should simply be virtual PC addresses (I will find a method
to handle multiple programs having the same virtual address spa
Hi,
X86 kernels version >2.6.32 (here 3.1.0) still gets the error "Cannot
allocate trampoline". Seems like the memory regions does not get
initialized correctly.
The stack trace is:
[0.00] Kernel panic - not syncing: Cannot allocate trampoline
[0.00]
[0.00] Pid: 0, comm: s
how much free memory do you have on host?
"free -m" will show that
On 12/15/11, Ankita (Garg) Goel wrote:
> I increased it to 4096MB, but still no luck. This has worked for me in the
> past, but not working now.
>
> On Thu, Dec 15, 2011 at 12:33 AM, biswabandan panda
> wrote:
>
>> increase the me
19 matches
Mail list logo
