Re: [sqlite] Problem compiling SQLite 3.8.4.1 with SQLITE_MAX_MMAP_SIZE=0

[Jan Nijtmans]

2014-03-14 12:06 GMT+01:00 Stephan Beal :
> i posted it to sqlite-dev on Feb. 19th, but Brad was on CC, and maybe that
> list elides mails with CCs.

Well, I just finished running the test-suite "make test" for the Cygwin64
build of SQLite 3.8.4.1 (with the SQLITE_MAX_MMAP_SIZE=0 fix).

For people interested, the following 7 tests had to be excluded
because they crash somehow (still to be investigated why):
test/multiplex.test
test/multiplex2.test
test/multiplex3.test
test/pager1.test
test/pager4.test
test/wal2.test
test/walnoshm.test

The previous test-suite run (SQLite 3.8.3.1, with the same 7 exclusions) had the
following result:
104 errors out of 132885 tests

Now, I got:
89 errors out of 133018 tests
Failures on these tests: backup2-6 crash8-5.1.2 crash8-5.2.2
crash8-5.3.2 crash8-5.4.2 crash8-5.5.2 crash8-5.6.2 crash8-5.7.2
crash8-5.8.2 crash8-5.9.2 delete-8.1 delete-8.2 delete-8.3 delete-8.4
delete-8.5 delete-8.6 filectrl-1.5 incrblob3-7.2 journal2-1.1
journal2-1.2 journal2-1.20 journal2-2.2 journal2-2.4 memsubsys1-4.5
memsubsys1-7.5 quota-2.1.2.1 quota-2.1.4 quota-2.1.5 quota-2.2.2
quota-2.2.3 quota-2.4.1 quota-2.4.2 quota-2.4.3 quota-2.4.4
quota-2.4.99 quota-3.1.4 quota-3.1.5 quota-3.2.6 quota-3.2.7
quota-3.2.8 quota-3.2.9 quota-3.3.1 quota-4.1.7 quota-4.1.8
quota-4.1.10 quota-4.1.11 quota-4.4.4 quota-4.4.6 quota-4.4.7
quota2-1.13 quota2-1.15 quota2-1.16 sync-1.1 sync-1.2 sync-1.3
temptable-6.2 temptable-6.3 temptable-6.6 tkt-f3e5abed55-2.5 uri-1.3.2
uri-1.3.4 uri-1.4.2 uri-1.4.4 uri-2.1.2 uri-2.2.2 uri-2.5.2 uri-2.6.2
uri-2.7.2 uri-2.8.2 uri-2.9.2 uri-2.10.2 uri-2.12.2 wal3-5.3 wal3-5.5
wal3-6.1.4 wal3-6.1.5 wal3-6.1.6 wal3-6.2.3 wal3-6.2.4 wal3-6.2.5
wal3-7.1.3 wal3-7.1.4 wal3-7.2.1 wal3-7.2.2 wal3-8.6 wal3-8.6
zerodamage-2.0 zerodamage-2.1 zerodamage-3.1
All memory allocations freed - no leaks
Memory used:  now  0  max 126424  max-size   33554304
Allocation count: now  0  max1507718
Page-cache used:  now  0  max 13  max-size  65784
Page-cache overflow:  now  0  max7639856
Scratch memory used:  now  0  max  0
Scratch overflow: now  0  max  74752  max-size  66540

That's 133 test-cases more, but 15 less failing test-cases!
That's progress! Give that the Cygwin VFS is built on top of the win32 API, and
that I have McAfee running (which tries to interfere too!) that's not
bad at all!

Regards,
Jan Nijtmans
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] Problem compiling SQLite 3.8.4.1 with SQLITE_MAX_MMAP_SIZE=0

[Stephan Beal]

On Fri, Mar 14, 2014 at 12:04 PM, Jan Nijtmans wrote:

> Hey, that's exactly the same solution! Pity that Brad didn't report
> that upstream, the fix could have been in SQLite 3.8.4.1 already.
>

i posted it to sqlite-dev on Feb. 19th, but Brad was on CC, and maybe that
list elides mails with CCs.

-- 
- stephan beal
http://wanderinghorse.net/home/stephan/
http://gplus.to/sgbeal
"Freedom is sloppy. But since tyranny's the only guaranteed byproduct of
those who insist on a perfect world, freedom will have to do." -- Bigby Wolf
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] Problem compiling SQLite 3.8.4.1 with SQLITE_MAX_MMAP_SIZE=0

[Jan Nijtmans]

2014-03-14 11:48 GMT+01:00 Stephan Beal :
> See also this fix:
>
> http://fossil.wanderinghorse.net/repos/libfossil/index.cgi/info/a9263730f2
>
> that's from Brad Harder, for BSD.

Hey, that's exactly the same solution! Pity that Brad didn't report
that upstream, the fix could have been in SQLite 3.8.4.1 already.

Thanks!

Regards,
 Jan Nijtmans
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users

Re: [sqlite] Problem compiling SQLite 3.8.4.1 with SQLITE_MAX_MMAP_SIZE=0

[Stephan Beal]

On Fri, Mar 14, 2014 at 11:46 AM, Jan Nijtmans wrote:

> I just found out that I was too quick upgrading libfossil to use
> the latest SQLite 3.8.4.1 amalgamation:
>   <
> http://fossil.wanderinghorse.net/repos/libfossil/index.cgi/info/2de81ae7b6
> >
> this broke the Cygwin build  ;-(. The reason is that libfossil is very
> strict, considering any warning as error, otherwise I wouldn't even
> have noticed this.
> CC [obj/sqlite3.o] ...
> src/sqlite3.c:52986:20: error: 'sqlite3BtreeSetMmapLimit' defined
> but not used [-Werror=unused-function]
>  SQLITE_PRIVATE int sqlite3BtreeSetMmapLimit(Btree *p,
> sqlite3_int64 szMmap){
>


See also this fix:

http://fossil.wanderinghorse.net/repos/libfossil/index.cgi/info/a9263730f2

that's from Brad Harder, for BSD.

-- 
- stephan beal
http://wanderinghorse.net/home/stephan/
http://gplus.to/sgbeal
"Freedom is sloppy. But since tyranny's the only guaranteed byproduct of
those who insist on a perfect world, freedom will have to do." -- Bigby Wolf
___
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users