compiler errors

compiler errors

[satadru pramanik]

I was compiling glib-1.1.2, and came up with an internal compiler error.
I have glibc .961212-1o

Should I complain, or move on to glibc2?
Errors:
testglib.c:966: internal error--unrecognizable insn:
(insn 9133 9130 9134 (set (mem:SI (plus:SI (reg:SI 16 r16)
                (const_int 65536)))
        (reg:SI 17 r17)) -1 (insn_list 5710 (insn_list 9130 (insn_list:REG_DEP_OUTPUT 5741 (insn_list:REG_DEP_OUTPUT 5713 (insn_list:REG_DEP_OUTPUT 5729 (insn_list:REG_DEP_OUTPUT 5736 (insn_list:REG_DEP_OUTPUT 5752 (insn_list:REG_DEP_OUTPUT 5764 (insn_list:REG_DEP_OUTPUT 5759 (insn_list:REG_DEP_OUTPUT 5772 (insn_list:REG_DEP_OUTPUT 5784 (insn_list:REG_DEP_OUTPUT 5779 (insn_list:REG_DEP_OUTPUT 5799 (insn_list:REG_DEP_OUTPUT 9126 (insn_list:REG_DEP_OUTPUT 9129 (nil))))))))))))))))
    (nil))
toplev.c:1360: Internal compiler error in function fatal_insn
make: *** [testglib.o] Error 1
[user@real glib-1.1.12]# 


[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Re: compiler errors

[David Edelsohn]

	You do not mention which release of EGCS you are using.  This may
have been fixed in egcs-1.1.1, but you need to use Gary Thomas's special
build of egcs-1.1.1 unless your system is running glibc2.1.

David

[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Re: compiler errors

[satadru pramanik]

egcs 1.1.1c
cheers,
sat.


On Wed, 13 Jan 1999, David Edelsohn wrote:

	You do not mention which release of EGCS you are using.  This may
have been fixed in egcs-1.1.1, but you need to use Gary Thomas's special
build of egcs-1.1.1 unless your system is running glibc2.1.

David



----
satadru pramanik `satadru@umich.edu'
"Put your hand on a hot stove for a minute, and it seems like an hour. Sit
with a pretty girl for an hour, and it seems like a minute. THAT'S
relativity." 
-Albert Einstein
---


[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

Re: compiler errors

[David Edelsohn]

>>>>> satadru pramanik writes:

satadru> egcs 1.1.1c

	Hmm, well I thought that problem had been addressed in egcs-1.1.1.
The problem is that GCC is generating a memory reference with a
displacement (65536) which is larger than 16 bits.  Someone needs to try
to track down why the PowerPC port thinks that is a valid address instead
of splitting it into an intermediate offset in a register.

	Either GCC is generating that offset itself and the normal tests
for valid types of addressing are failing.  Or, something in the PowerPC
port explicitly is generating a memory operation which bypasses the tests
and is not explicitly checking for exceeding this offset limit.

David

[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]