Mailing Lists

Sunday, 12 June 2016

[gccsdk] [Bug 262] New: GCC 4.7.4 Rel 2 buffer overflow detected when -Os used

http://www.riscos.info/bugzilla3/show_bug.cgi?id=262

Summary: GCC 4.7.4 Rel 2 buffer overflow detected when -Os used
Product: GCC/GCCSDK
Version: other
Platform: Other
OS/Version: RISC OS
Status: NEW
Severity: feature request
Priority: P1
Component: C compiler
AssignedTo: leenoar@sky.com
ReportedBy: dave@davespace.co.uk
Estimated Hours: 0.0

Created attachment 86
--> http://www.riscos.info/bugzilla3/attachment.cgi?id=86
Preprocessed source file to repro the issue.

GCCSDK 4.7.4 release 2 built locally on Ubuntu 14.04.4 LTS.

Compiler crashes when -Os is used. Log attached. Any other -O level seems fine.

Preprocessed source attached.

jenkins@microserver:~/workspace/PrivateEye/libs/appengine$
/home/dave/src/gccsdk/cross/bin/arm-unknown-riscos-gcc -c walk-cont.pp.c -o
datastruct/hash/walk-cont.o -Os
*** buffer overflow detected ***:
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1 terminated
======= Backtrace: =========
/lib/x86_64-linux-gnu/libc.so.6(+0x7329f)[0x7f848493b29f]
/lib/x86_64-linux-gnu/libc.so.6(__fortify_fail+0x5c)[0x7f84849d2bbc]
/lib/x86_64-linux-gnu/libc.so.6(+0x109a90)[0x7f84849d1a90]
/lib/x86_64-linux-gnu/libc.so.6(+0x108c2b)[0x7f84849d0c2b]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1[0x9ef503]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(final_scan_insn+0xba9)[0x625b39]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(final+0x24a)[0x6264ea]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1[0x6266e6]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(execute_one_pass+0xfe)[0x705b9e]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(execute_pass_list+0x15)[0x705ee5]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(execute_pass_list+0x27)[0x705ef7]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(execute_pass_list+0x27)[0x705ef7]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(tree_rest_of_compilation+0x73)[0x7da763]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1[0x59644a]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(cgraph_optimize+0x6bc)[0x597d3c]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(cgraph_finalize_compilation_unit+0x75)[0x5980f5]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(c_write_global_declarations+0xf8)[0x4de978]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1(toplev_main+0xacc)[0x79516c]
/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5)[0x7f84848e9f45]
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1[0x4cde5f]
======= Memory map: ========
00400000-010fd000 r-xp 00000000 08:42 6678470
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1
012fc000-012fd000 r--p 00cfc000 08:42 6678470
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1
012fd000-0130d000 rw-p 00cfd000 08:42 6678470
/home/dave/src/gccsdk/cross/libexec/gcc/arm-unknown-riscos/4.7.4/cc1
0130d000-014a4000 rw-p 00000000 00:00 0
026ae000-02938000 rw-p 00000000 00:00 0 [heap]
7f84848c8000-7f8484a82000 r-xp 00000000 08:42 2101535
/lib/x86_64-linux-gnu/libc-2.19.so
7f8484a82000-7f8484c82000 ---p 001ba000 08:42 2101535
/lib/x86_64-linux-gnu/libc-2.19.so
7f8484c82000-7f8484c86000 r--p 001ba000 08:42 2101535
/lib/x86_64-linux-gnu/libc-2.19.so
7f8484c86000-7f8484c88000 rw-p 001be000 08:42 2101535
/lib/x86_64-linux-gnu/libc-2.19.so
7f8484c88000-7f8484c8d000 rw-p 00000000 00:00 0
7f8484c8d000-7f8484ca3000 r-xp 00000000 08:42 2110253
/lib/x86_64-linux-gnu/libgcc_s.so.1
7f8484ca3000-7f8484ea2000 ---p 00016000 08:42 2110253
/lib/x86_64-linux-gnu/libgcc_s.so.1
7f8484ea2000-7f8484ea3000 rw-p 00015000 08:42 2110253
/lib/x86_64-linux-gnu/libgcc_s.so.1
7f8484ea3000-7f8484f89000 r-xp 00000000 08:42 3145839
/usr/lib/x86_64-linux-gnu/libstdc++.so.6.0.19
7f8484f89000-7f8485188000 ---p 000e6000 08:42 3145839
/usr/lib/x86_64-linux-gnu/libstdc++.so.6.0.19
7f8485188000-7f8485190000 r--p 000e5000 08:42 3145839
/usr/lib/x86_64-linux-gnu/libstdc++.so.6.0.19
7f8485190000-7f8485192000 rw-p 000ed000 08:42 3145839
/usr/lib/x86_64-linux-gnu/libstdc++.so.6.0.19
7f8485192000-7f84851a7000 rw-p 00000000 00:00 0
7f84851a7000-7f84851aa000 r-xp 00000000 08:42 2097260
/lib/x86_64-linux-gnu/libdl-2.19.so
7f84851aa000-7f84853a9000 ---p 00003000 08:42 2097260
/lib/x86_64-linux-gnu/libdl-2.19.so
7f84853a9000-7f84853aa000 r--p 00002000 08:42 2097260
/lib/x86_64-linux-gnu/libdl-2.19.so
7f84853aa000-7f84853ab000 rw-p 00003000 08:42 2097260
/lib/x86_64-linux-gnu/libdl-2.19.so
7f84853ab000-7f84854b0000 r-xp 00000000 08:42 2097255
/lib/x86_64-linux-gnu/libm-2.19.so
7f84854b0000-7f84856af000 ---p 00105000 08:42 2097255
/lib/x86_64-linux-gnu/libm-2.19.so
7f84856af000-7f84856b0000 r--p 00104000 08:42 2097255
/lib/x86_64-linux-gnu/libm-2.19.so
7f84856b0000-7f84856b1000 rw-p 00105000 08:42 2097255
/lib/x86_64-linux-gnu/libm-2.19.so
7f84856b1000-7f84856d4000 r-xp 00000000 08:42 2101493
/lib/x86_64-linux-gnu/ld-2.19.so
7f84858a8000-7f84858d1000 rw-p 00000000 00:00 0
7f84858d1000-7f84858d3000 rw-p 00000000 00:00 0
7f84858d3000-7f84858d4000 r--p 00022000 08:42 2101493
/lib/x86_64-linux-gnu/ld-2.19.so
7f84858d4000-7f84858d5000 rw-p 00023000 08:42 2101493
/lib/x86_64-linux-gnu/ld-2.19.so
7f84858d5000-7f84858d6000 rw-p 00000000 00:00 0
7ffec4801000-7ffec4822000 rw-p 00000000 00:00 0
[stack]
7ffec4901000-7ffec4903000 r-xp 00000000 00:00 0 [vdso]
ffffffffff600000-ffffffffff601000 r-xp 00000000 00:00 0
[vsyscall]
datastruct/hash/walk-cont.c: In function 'hash_walk_continuation':
datastruct/hash/walk-cont.c:95:1: internal compiler error: Aborted
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gccsdk.riscos.info/> for instructions.

--
Configure bugmail: http://www.riscos.info/bugzilla3/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.

_______________________________________________
GCCSDK mailing list gcc@gccsdk.riscos.info
Bugzilla: http://www.riscos.info/bugzilla/index.cgi
List Info: http://www.riscos.info/mailman/listinfo/gcc
Main Page: http://www.riscos.info/index.php/GCCSDK

Wednesday, 8 June 2016

Re: [Rpcemu] Pulseaudio has constant CPU load even when sound is disabled

In article
<CAOnWdogzYmCLi2FN2NfTtp54EwNe8H00V4HMwQEkUbMnZaH3HA@mail.gmail.com>,
Reuben Thomas <rrt@sc3d.org> wrote:
> Using rpcemu 0.8.14, emulating an SA-RPC running RISC OS 3.71, I notice
> that pulseaudio is running with a constant appreciable load (~20% of a
> core on my machine), which stops when I quit rpcemu.

> This would be less than ideal even if I had sound switched on, but it's
> off!

I can't advise on the reasons why the above occurs. TBH one of the first
things I do with any new Linux install is to try and beat Pulseaudio to
death with a big stick as I find it a real PITA! If I can't kill it, I do
what I can to isolate and bypass it.

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

[Rpcemu] Running 1280x1024

[Apologies for not replying properly; I had list delivery switched off for the last few years, and only switched it back on after getting a reply to my message, which I therefore didn't receive, but read in the archives.]

Thanks, Peter Howkins.

As I don't seem to have a proper 3.7x !Boot, and am only using the emulator lightly, I'll wait until the next rpcemu release to use larger screen sizes (or until I find time to update my !Boot), and to re-enable dynarec (though on my modest 2011-era 4-core Intel Core i5 system, performance is perfectly acceptable even without).

http://rrt.sc3d.org

[Rpcemu] Pulseaudio has constant CPU load even when sound is disabled

Using rpcemu 0.8.14, emulating an SA-RPC running RISC OS 3.71, I notice that pulseaudio is running with a constant appreciable load (~20% of a core on my machine), which stops when I quit rpcemu.

This would be less than ideal even if I had sound switched on, but it's off!

http://rrt.sc3d.org

Monday, 6 June 2016

Re: [Rpcemu] Running 1280x1024

On Mon, Jun 06, 2016 at 01:01:03AM +0100, Reuben Thomas wrote:
> I'm running rpcemu 0.8.14 on an x86_64 GNU/Linux system (Ubuntu 14.04).
> I am using RISC OS 3.71, with StrongARM emulation, 2Mb VRAM selected, and
> 32Mb of RAM. I am not using dynrec (I found it unusably unstable) and I
> have "reduce CPU usage" selected.

The dynarec issue and the one you mention below are probably the same
thing ...

> I am running happily in 1024x768x256 colours. I have the (apparently
> default) AKF85 monitor selection.
> When I switch to 1280x1024, the display change normally works fine, but I
> get constant crashes until I switch back to 1024x768.
> Any clues?
> If I've failed to supply enough/the right information, do let me know!

The last version added auto patching for 8MB VRAM for RISC OS 3.7 if 2MB
VRAM is sslected. There's a bug in it (now fixed in the development
repository).

Basically to enable large screen modes in 3.7 select 0MB VRAM and find the
BandLimit line in the file !Boot.RO370Hook.Boot.PreDesk.Configure.!Run add
a 0 to the end of each of those numbers, this should allow larger
modes with 0MB VRAM (don't do this on a real machine, but it's fine on an
emulator).

Peter

--
Peter Howkins
peter.howkins@marutan.net

_______________________________________________
Rpcemu mailing list
Rpcemu@riscos.info
http://www.riscos.info/cgi-bin/mailman/listinfo/rpcemu

Sunday, 5 June 2016

[Rpcemu] Running 1280x1024

I'm running rpcemu 0.8.14 on an x86_64 GNU/Linux system (Ubuntu 14.04).

I am using RISC OS 3.71, with StrongARM emulation, 2Mb VRAM selected, and 32Mb of RAM. I am not using dynrec (I found it unusably unstable) and I have "reduce CPU usage" selected.

I am running happily in 1024x768x256 colours. I have the (apparently default) AKF85 monitor selection.

When I switch to 1280x1024, the display change normally works fine, but I get constant crashes until I switch back to 1024x768.

Any clues?

If I've failed to supply enough/the right information, do let me know!

http://rrt.sc3d.org

[Rpcemu] Web site typos

On the home page:

"but not limited too" → "but not limited to"

"How to get code committed into the repository" should probably be a heading.

http://rrt.sc3d.org