[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: make handling of prerequisites that interlock?
From: |
Ben Pfaff |
Subject: |
Re: make handling of prerequisites that interlock? |
Date: |
Tue, 05 Apr 2005 09:05:34 -0700 |
User-agent: |
Gnus/5.1007 (Gnus v5.10.7) Emacs/21.4 (gnu/linux) |
Ted Stern <address@hidden> writes:
> On 4 Apr 2005 at 13:56 UTC-0700, Ben Pfaff wrote:
>> I'm having some trouble with a set of dependencies that
>> indirectly depend on a common file. Here's a minimal makefile
>> that demonstrates the problem:
[...]
>> # Turn off implicit rules to make `make -d' output easier to read.
>> .SUFFIXES:
>> %: %,v
>> %: RCS/%,v
>> %: RCS/%
>> %: s.%
>> %: SCCS/s.%
>
> This is a bit of overkill. I would recommend instead to always use
> the -R command line flag, which turns off all built-in rules and
> variables. It is sort of like declaring IMPLICIT NONE in Fortran ;-).
This is simply part of my stripped-down example. It is not
anything I would use in practice. (However, I was not aware of
-R before.)
>> ----------------------------------------------------------------------
>>
>> When I run this in a tight loop, as with
>>
>> rm -f foo
>> touch x d
>> while ! grep newer foo; do
>> make clean > /dev/null
>> make > /dev/null
>> echo x > x
>> make -d > foo
>> done
>>
>> it typically iterates for a few seconds (with most iterations not
>> seeing any change in file times, presumably because everything
>> happens too fast), and then stops with this output in foo:
>
> I can bet you're going to have problems here! Why are you doing this?
To reliably demonstrate a problem that I have seen elsewhere in
other circumstances.
> I'd also recommend trying a different filesystem. Reiserfs, for
> example, handles small files better.
Why do you think this issue is related to small file handling?
--
"Debian for hackers, Red Hat for suits, Slackware for loons."
--CmdrTaco <URL:http://slashdot.org/articles/99/03/22/0928207.shtml>
Re: make handling of prerequisites that interlock?, Boris Kolpackov, 2005/04/06