| 1 |
Joachim wrote: |
| 2 |
> Joachim pisze: |
| 3 |
>> Duncan pisze: |
| 4 |
>>> Here it is, from your attached changelog (thanks): |
| 5 |
>>> /usr/qt/3/bin/uic -nounload -tr tr2i18n -i kshortcutdialog_simple.h |
| 6 |
>>> ./kshortcutdialog_simple.ui > kshortcutdialog_simple.cpp.temp ; |
| 7 |
>>> ret=$?; |
| 8 |
>>> [snip] |
| 9 |
>>> |
| 10 |
>>> /bin/sh: line 3: 4547 Segmentation fault /usr/qt/3/bin/uic |
| 11 |
>>> -nounload -tr tr2i18n -i kshortcutdialog_simple.h |
| 12 |
>>> ./kshortcutdialog_simple.ui > kshortcutdialog_simple.cpp.temp |
| 13 |
>>> echo '#include <klocale.h>' >> kshortcutdialog_advanced.cpp |
| 14 |
>>> make[3]: *** [kshortcutdialog_simple.cpp] Error 139 |
| 15 |
>>> make[3]: *** Waiting for unfinished jobs.... |
| 16 |
> I sent before I finished writing. Here is what I wanted to write: |
| 17 |
> |
| 18 |
> I had problems with uic segmentation faults because I had march=native |
| 19 |
> in my C/CXX FLAGS. When I changing march=native to march=k8 ( as in |
| 20 |
> http://en.gentoo-wiki.com/wiki/Safe_Cflags/AMD ) and reemerging qt-3 and |
| 21 |
> kdelibs solved my problem. |
| 22 |
> |
| 23 |
> Also I got an explanation ( |
| 24 |
> http://forums.gentoo.org/viewtopic-p-5795915.html#5795915 ): |
| 25 |
> qt3 is stripping flags. Some that must be compiled with the same flags |
| 26 |
> as qt don't strip flags. So the are compiled with different flags. So |
| 27 |
> they don't work properly. |
| 28 |
> |
| 29 |
|
| 30 |
|
| 31 |
Thank you, Joachim, for your insights. Having no experience with |
| 32 |
changing CFLAGS and CXXFLAGS, I should proceed cautiously. |
| 33 |
Doing "locate make.conf" generates a long list of such files in various |
| 34 |
directories. Guessing that /etc/make.conf may be the pertinent one, see |
| 35 |
that its first 6 lines read as follows: |
| 36 |
|
| 37 |
# These settings were set by the catalyst build script that |
| 38 |
automatically built this stage |
| 39 |
# Please consult /etc/make.conf.example for a more detailed example |
| 40 |
CFLAGS="-Os -march=opteron -mtune=opteron -fomit-frame-pointer -ggdb" |
| 41 |
CHOST="x86_64-pc-linux-gnu" |
| 42 |
CXXFLAGS="${CFLAGS}" |
| 43 |
|
| 44 |
The question now, I suppose, is whether -march=opteron is "safe" for my |
| 45 |
cpu. My /proc/cpuinfo file contains the following lines |
| 46 |
|
| 47 |
vendor_id : AuthenticAMD |
| 48 |
cpu family : 15 |
| 49 |
model : 65 |
| 50 |
model name : Dual-Core AMD Opteron(tm) Processor 2210 |
| 51 |
stepping : 2 |
| 52 |
|
| 53 |
The "Safe Cflags/AMD" web page you cited has a section on "22xx |
| 54 |
Dual-Core Opteron" consisting of 3 boxes. The first box consists of the |
| 55 |
following lines: |
| 56 |
|
| 57 |
vendor_id : AuthenticAMD |
| 58 |
cpu family : 15 |
| 59 |
model : 65 |
| 60 |
model name : Dual-Core AMD Opteron(tm) Processor 22xx |
| 61 |
stepping : 2 |
| 62 |
|
| 63 |
That closely matches my cpuinfo. |
| 64 |
|
| 65 |
The second box consists of these lines: |
| 66 |
|
| 67 |
CHOST="x86_64-pc-linux-gnu" |
| 68 |
CFLAGS="-march=opteron -O2 -pipe" |
| 69 |
CXXFLAGS="${CFLAGS}" |
| 70 |
|
| 71 |
Thus my -march=opteron setting seems "safe." Are the discrepancies in |
| 72 |
other parts of the CFLAGS line significant for my kdelibs problem? |
| 73 |
|
| 74 |
Is there anything in the third box under "22xx Dual-Core Opteron" that |
| 75 |
needs attention? |
| 76 |
|
| 77 |
-John |
| 78 |
|
| 79 |
-- |
| 80 |
John P. Burkett |
| 81 |
Department of Economics |
| 82 |
University of Rhode Island |
| 83 |
Kingston, RI 02881-0808 |
| 84 |
USA |
| 85 |
|
| 86 |
phone (401) 874-9195 |
Archives