Test results: egcs-971225 on i586-pc-linux-gnu(RH 5.0)

Test results: egcs-971225 on i586-pc-linux-gnu(RH 5.0)

[Byeong-ryeol Kim]

1. kernel 2.1.76, glibc-2.0.5c-13, binutils-2.8.1.0.18, libg++-2.8.0.b6.5,
  egcs-1.0.1-prerelease on Pentium MMX 166 with 64M EDO RAM(ASUS
  P/I-P55T2P4 rev 3.0 mainboard)

2. cd /usr/X11R6/src/egcs-971225 ; patch -p0 < README.libg++

3. sh configure --prefix=/usr --enable-shared --enable-threads=posix \
                --with-gnu-as --with-gnu-ld

4. make CFLAGS="-O2 -pipe -s" LDFLAGS="-s" bootstrap-lean

5. make install ; ldconfig

6. export TOPDIR=/usr/X11R6/src/egcs-971225
   export LIBRARYP_PATH=$TOPDIR/libraries/libio:$TOPDIR/libraries/ \
          libstdc++:$TOPDIR/libraries/libg++:/usr/lib
   export LD_LIBARY_PATH=$LIBRARY_PATH
   export LD_RUN_PATH=$LIBRARY_PATH

7. make -k check

8. Results:

                === libio Summary ===

# of expected passes            40

                === libstdc++ Summary ===

# of expected passes            24
# of unexpected failures        2
# of expected failures          4
...
FAIL: tstring.cc compilation
XFAIL: tstring.cc execution
XFAIL: tstring.cc output
FAIL: tstring.cc -O compilation
XFAIL: tstring.cc -O execution
XFAIL: tstring.cc -O output
...

                === gcc Summary ===

# of expected passes            5969
# of expected failures          6
# of unsupported tests          7

...
XFAIL: gcc.c-torture/compile/961203-1.c
XFAIL: gcc.failure/940409-1.c,  -O0  
XFAIL: gcc.failure/940409-1.c,  -O1  
XFAIL: gcc.failure/940409-1.c,  -O2  
XFAIL: gcc.failure/940409-1.c,  -O2 -fomit-frame-pointer
-finline-functions  
XFAIL: gcc.failure/940409-1.c,  -O2 -g  
...

                === g++ Summary ===

# of expected passes            3462
# of unexpected failures        1
# of unexpected successes       2
# of expected failures          79
# of untested testcases         6

...
XFAIL: g++.brendan/bit-fields2.C .* , (test for errors, line 11)
XFAIL: g++.bugs/900213_03.C missed by g++, warned by cfront, (test for
errors, line 24)
XFAIL: g++.bugs/900215_02.C , (test for errors, line 44)
XFAIL: g++.bugs/900321_01.C , (test for errors, line 22)
XFAIL: g++.bugs/900321_01.C , (test for errors, line 23)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 25)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 29)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 35)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 39)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 43)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 45)
XFAIL: g++.bugs/900322_01.C , (test for errors, line 49)
XFAIL: g++.bugs/900402_01.C  (test for errors, line 12)
XFAIL: g++.bugs/900404_04.C , (test for errors, line 18)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 26)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 27)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 28)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 30)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 31)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 32)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 34)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 35)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 36)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 38)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 39)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 40)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 42)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 43)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 44)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 46)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 47)
XFAIL: g++.bugs/900428_01.C , (test for errors, line 48)
XFAIL: g++.bugs/900519_05.C  (test for bogus messages, line 10)
XFAIL: g++.bugs/900519_05.C  (test for bogus messages, line 16)
XFAIL: g++.bugs/900519_05.C  Execution test
XFAIL: g++.jason/access7.C - wrongly ambiguous static member call, (test
for bogus messages, line 12)
XFAIL: g++.jason/ambig3.C - late parsing (test for bogus messages, line 12)
XFAIL: g++.jason/ambig3.C - late parsing (test for bogus messages, line 14)
XFAIL: g++.jason/cond.C , (test for errors, line 9)
XFAIL: g++.jason/cond.C , (test for errors, line 11)
XFAIL: g++.jason/cond.C , (test for errors, line 16)
XFAIL: g++.jason/opeq5.C - Execution test
XFAIL: g++.jason/pmf2.C failed conversion to method pointer (test for
errors, line 12)
XFAIL: g++.jason/ref4.C - references to functions (test for bogus
messages, line 4)
XFAIL: g++.jason/rfg16.C  (test for bogus messages, line 4)
XFAIL: g++.jason/rfg16.C  (test for bogus messages, line 5)
XFAIL: g++.jason/template18.C , (test for errors, line 9)
XFAIL: g++.law/arg11.C (test for excess errors)
XFAIL: g++.law/arm14.C .* , (test for errors, line 21)
XFAIL: g++.law/arm2.C (test for excess errors)
XFAIL: g++.law/array2.C (test for excess errors)
XFAIL: g++.law/code-gen3.C - Execution test
XFAIL: g++.law/enum5.C (test for excess errors)
XFAIL: g++.law/friend4.C (test for excess errors)
XFAIL: g++.law/inline2.C (test for excess errors)
XFAIL: g++.law/missed-error1.C , (test for errors, line 10)
XFAIL: g++.law/operators29.C (test for excess errors)
XFAIL: g++.law/union1.C (test for excess errors)
XFAIL: g++.law/visibility22.C (test for excess errors)
XFAIL: g++.law/visibility23.C .* , (test for errors, line 26)
XFAIL: g++.law/visibility23.C (test for excess errors)
XFAIL: g++.law/vtable1.C (test for excess errors)
XFAIL: g++.mike/misc5.C uses same name 9.3 , (test for errors, line 10)
XFAIL: g++.mike/ns10.C (test for excess errors)
XFAIL: g++.mike/ns10.C  Execution test
XFAIL: g++.mike/ns11.C (test for excess errors)
XFAIL: g++.mike/ns11.C  Execution test
XFAIL: g++.mike/ns13.C (test for excess errors)
XFAIL: g++.mike/ns14.C (test for excess errors)
XFAIL: g++.mike/ns3.C  (test for errors, line 1)
XFAIL: g++.mike/ns5.C  (test for errors, line 5)
XFAIL: g++.mike/ns7.C  (test for errors, line 8)
XFAIL: g++.mike/ns9.C (test for excess errors)
XFAIL: g++.mike/ns9.C  Execution test
XFAIL: g++.mike/p1989.C - (test for bogus messages, line 339)
XFAIL: g++.mike/p6058.C  (test for errors, line 7)
FAIL: g++.mike/p6610a.C  Execution test
XFAIL: g++.mike/p7325.C *-*-* Execution test
XFAIL: g++.niklas/t135.C (test for excess errors)
UNTESTED: g++.other/dll-1.C: Test skipped: Skip if not target: arm-*pe
UNTESTED: g++.other/dll-2.C: Test skipped: Skip if not target: arm-*pe
UNTESTED: g++.other/dll-3.C: Test skipped: Skip if not target: arm-*pe
UNTESTED: g++.other/dll-4.C: Test skipped: Skip if not target: arm-*pe
UNTESTED: g++.other/dll-5.C: Test skipped: Skip if not target: arm-*pe
UNTESTED: g++.other/dll-6.C: Test skipped: Skip if not target: arm-*pe
...

                === g77 Summary ===

# of expected passes            130
# of unexpected failures        2

...
FAIL: g77.f-torture/execute/dnrm2.f execution,  -O2 -fomit-frame-pointer
-finline-functions -funroll-loops
FAIL: g77.f-torture/execute/dnrm2.f execution,  -O2 -fomit-frame-pointer
-finline-functions -funroll-all-loops
....

 "Where there is a will, there is a way."  jinbo21@soback.kornet.nm.kr 
                                           kbeyl@kids.kotel.co.kr      
  For the future of you and me!            hitel: jinbo21

Re: linux kernel-2.1.76 panic(..egcs-971225 on i586-pc-linux-gnu..)

[Byeong-ryeol Kim]

On Tue, 30 Dec 1997, Byeong-ryeol Kim wrote:

> 1. kernel 2.1.76, glibc-2.0.5c-13, binutils-2.8.1.0.18, libg++-2.8.0.b6.5,
>   egcs-1.0.1-prerelease on Pentium MMX 166 with 64M EDO RAM(ASUS
>   P/I-P55T2P4 rev 3.0 mainboard)
...
 I compiled egcs-971225 with vanilla glibc-2.0.6 on RH 5.0 linux 
system, and the result of 'make -k check' was the same as with
glibc-2.0.5c-13.
 I tried to compile several applications and kernel-2.1.76.
 It was successful in most case.
 But, as for kernel-2.1.76, I could not boot the compiled kernel.
 The log file says as follows:
....
gcc -D__KERNEL__ -I/usr/X11R6/src/linux/include -Wall -Wstrict-prototypes
-O2 -f
omit-frame-pointer -pipe -fno-strength-reduce -mcpu=pentium -march=pentium
-mali
gn-loops=2 -malign-jumps=2 -malign-functions=2   -c -o init/main.o
init/main.c
/usr/X11R6/src/linux/include/asm/string.h: In function
`__constant_memcpy':
In file included from /usr/X11R6/src/linux/include/linux/string.h:39,
                 from /usr/X11R6/src/linux/include/linux/signal.h:64,
                 from /usr/X11R6/src/linux/include/linux/sched.h:22,
                 from init/main.c:17:
/usr/X11R6/src/linux/include/asm/string.h:448: warning: control reaches
end of n
on-void function
/usr/X11R6/src/linux/include/asm/string.h: In function
`__constant_c_and_count_m
emset':
/usr/X11R6/src/linux/include/asm/string.h:599: warning: control reaches
end of n
on-void function
.... 
 The same warnings continually appeared, and after compiling and
installing the kernel and modules, I stucked at the beginning of 
kernel booting. It seems to be related with kswapd, I partially 
read from OOPS message.
 Sorry, I could not log OOPS, this was complete lock-up of the 
system.

 BTW, I remember that I could boot some kernels compiled with egcs
 before, such as 2.0.33, 2.1.7x. Though, startx failed on 2.0.33, 
as stated in egcs-FAQ. This case with egcs-971225 is different. 

 "Where there is a will, there is a way."  jinbo21@soback.kornet.nm.kr 
                                           kbeyl@kids.kotel.co.kr      
  For the future of you and me!            hitel: jinbo21

Re: linux kernel-2.1.76 panic(..egcs-971225 on i586-pc-linux-gnu..)

[Byeong-ryeol Kim]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1796 bytes --]

On Wed, 31 Dec 1997, Byeong-ryeol Kim wrote:
...
>  But, as for kernel-2.1.76, I could not boot the compiled kernel.
...
 In short, the boot problem of kernel 2.1.76 was proved that it
is not directly related with egcs.
 I should have reported it to the linux-kernel group.

 At first, I though it was related with egcs's optimization, as
discussed in this group.
 So I repeated to compile and try to boot kernel 2.1.76, 2.1.77 
with gcc-2.7.2.3-8, and  egcs-1.0.1-prerelease several times, but
the result was that though memcpy warning disappeared, I could not
boot compiled kernels.

 As a last resort, I changed kernel config file(/usr/src/linux/.config).
 I changed one option about X.25 LLC driver(CONFIG_LLC), and 
after compiling the kernel with gcc-2.7.2.3-8 of rh 5.0, the boot 
problem went away. 

 Though I do not use this driver, I usually compile kernels
with almost all modules except some(LLC, some SCSI modules...)
 When I report previous boot failure of kernel-2.1.76 compiled
with egcs-971225, I enabled CONFIG_LLC in .config file, and it
was the first time for me to compile LLC driver into kernel.
 It took 2 days for me to notice LLC driver was the culprit, sigh...

 As a conclusion, the boot failure of kernel 2.1.76 was not 
due to bugs of egcs-971225, though there might be some problems 
to compile and use linux kernels with egcs. 
 I'll retry to compile kernel-2.1.7x(without 'CONFIG_LLC' ¿É¼Ç) 
with egcs-1.0.1, egcs-971225, and report the results.

 I'm sorry for my incorect report that might make unfavorable
impression about egcs on linux users.

 "Where there is a will, there is a way."  jinbo21@soback.kornet.nm.kr 
                                           kbeyl@kids.kotel.co.kr      
  For the future of you and me!            hitel: jinbo21

Re: linux kernel-2.1.76 panic(..egcs-971225 on i586-pc-linux-gnu..)

[Byeong-ryeol Kim]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 927 bytes --]

On Sun, 4 Jan 1998, Byeong-ryeol Kim wrote:
...
>  As a conclusion, the boot failure of kernel 2.1.76 was not 
> due to bugs of egcs-971225, though there might be some problems 
> to compile and use linux kernels with egcs. 
>  I'll retry to compile kernel-2.1.7x(without 'CONFIG_LLC' ¿É¼Ç) 
> with egcs-1.0.1, egcs-971225, and report the results.
...
 I re-compiled egcs-1.0.1(egcs-2.90.23), kernel-2.1.77 with it, 
booted successfuly. (experimental) LCC code of kernel-2.1.77 is 
the same as that of kernel-2.1.76.
 I inserted "-march=pentium -malign-loops=2 -malign-jumps=4 
-malign-functions=4" in the CFLAGS of arch/i386/Makefile.

 I'll do the same work with egcs-971225 and report the result
tomorrow.
  
 "Where there is a will, there is a way."  jinbo21@soback.kornet.nm.kr 
                                           kbeyl@kids.kotel.co.kr      
  For the future of you and me!            hitel: jinbo21

Re: linux kernel-2.1.76 panic(..egcs-971225 on i586-pc-linux-gnu..)

[Jeffrey A Law]

  In message <Pine.LNX.3.96.971231192116.518H-100000@progress.plw.net>you write
  > On Tue, 30 Dec 1997, Byeong-ryeol Kim wrote:
  >  I compiled egcs-971225 with vanilla glibc-2.0.6 on RH 5.0 linux 
  > system, and the result of 'make -k check' was the same as with
  > glibc-2.0.5c-13.
  >  I tried to compile several applications and kernel-2.1.76.
  >  It was successful in most case.
  >  But, as for kernel-2.1.76, I could not boot the compiled kernel.
  >  The log file says as follows:
[ ... ]
I wonder if this is the ADDRESSOF optimization mis-compiling ioport.c

Can you try building ioport.c without optimization and see if your
kernel boots?

This would have worked with egcs-1.0.x since those releases do not
have the ADDRESSOF optimization.

It's also possible a bug in the egcs asm support is causing your
problems (I'm discussing it with some linux kernel folks).  I haven't
build a 2.1.* kernel in a long time, though I do have 2.0.33 kernels
built with egcs running on both my pcs.


jeff