This is the mail archive of the gcc-bugs@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: bootstrap comparison failure on alpha


> 
> >>>>> "Brad" == Brad Lucier <lucier@math.purdue.edu> writes:
> 
>     Brad> I don't have install privileges, and it still doesn't
>     Brad> bootstrap on alphaev6-unknown-linux-gnu.
> 
> OK.  Are you able to figure out why?

No, and I've tried.  For example, build.log reports that reload.o differ;
I've created two dummy directories stage1test and stage2test at the
same level as gcc in the build directory, copied reload.c to them,
generated reload.i with stage1/gcc and stage2/gcc (it's the same)
and compiled reload.i with stage1/cc1 and stage2/cc1 with the command
lines

../gcc/stage2/cc1 -g -O2 -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -Wtraditional -pedantic -Wno-long-long reload.i -da

../gcc/stage1/cc1 -g -O2 -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -Wtraditional -pedantic -Wno-long-long -da reload.i

All the report files are different; the diffs are at

http://www.math.purdue.edu/~lucier/diffs.gz

For example, the diff of reload.i.00.rtl begins

reload.i.00.rtl:
24327c24327
< (note 9 7 11 0x200007d75c0 NOTE_INSN_BLOCK_BEG 0)
---
> (note 9 7 11 0x200007d7cc0 NOTE_INSN_BLOCK_BEG 0)
25029c25029
< (note 252 251 253 0x200007d75c0 NOTE_INSN_BLOCK_END 0)
---
> (note 252 251 253 0x200007d7cc0 NOTE_INSN_BLOCK_END 0)
25137c25137
< (note 42 40 44 0x20000ae8f40 NOTE_INSN_BLOCK_BEG 0)
---
> (note 42 40 44 0x20000ae9640 NOTE_INSN_BLOCK_BEG 0)
25355c25355
< (note 126 125 128 0x200007e9e80 NOTE_INSN_BLOCK_BEG 0)
---
> (note 126 125 128 0x200007ec580 NOTE_INSN_BLOCK_BEG 0)
25499c25499
< (note 175 174 176 0x200007e9e80 NOTE_INSN_BLOCK_END 0)
---
> (note 175 174 176 0x200007ec580 NOTE_INSN_BLOCK_END 0)
25532c25532
< (note 186 185 188 0x200007eda80 NOTE_INSN_BLOCK_BEG 0)
---
> (note 186 185 188 0x200007f4180 NOTE_INSN_BLOCK_BEG 0)
25668c25668
< (note 232 231 233 0x200007eda80 NOTE_INSN_BLOCK_END 0)
---
> (note 232 231 233 0x200007f4180 NOTE_INSN_BLOCK_END 0)

Whether this is relevant or not, I cannot tell.  But the reload.s
files definitely differ using the two cc1's.

Brad

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]