Import SQLite 3.10.2.

Import SQLite 3.10.2.

Ran through the import script in third_party/sqlite/README.Chromium,
including the SQLite test suite.  No new errors observed.

SQLite changes are at http://www.sqlite.org/changes.html , Chromium
previously used 3.8.7.4.

All patches were applied and the results reviewed to make sure
backported patches were safe to remove, and retained patches were still
covering what was necessary.

Keep fts4 disabled, and also the new fts3 virtual table and unicode61
tokenizer.  Once enabled, these are very hard to disable, and there
doesn't seem to be any pressure to enable them.  Set
SQLITE_DEFAULT_PCACHE_INITSZ to 0 so SQLite does not preallocate 100
pages to pcache (both so Chromium can manage memory, also because our
code disables shared pcache).  Other SQLITE_* flags were reviewed for
applicability, none looked essential.

BUG=579743
R=michaeln@chromium.org

Committed: https://chromium.googlesource.com/chromium/src/+/e0b6a82cadecbda7e005921c2416fc7afb3542aa

[Scott Hess - ex-Googler, 2016-01-21 01:44:23]

shess@chromium.org changed reviewers:
+ michaeln@chromium.org

[Scott Hess - ex-Googler, 2016-01-21 01:44:26]

Posting to give you a chance to rubber-stamp LGTM it :-).

The trybots won't touch the amalgamation/sqlite3.c file, but I wrote a hack to
split it up and ran that through the trybots here:
   https://codereview.chromium.org/1560323002
That may differ in minor ways from this, I'll hack it through again before
landing.

Previously I tended to let the patches depart from perfect to prevent
distracting from whatever was _actually_ changing, but this time I let them
update to get the line numbers and stuff updated.  The substantive changes:
 - 0001-test-SQLite-tests-compiling-on-Linux.patch needed a linking change for
sqlite3_analyzer which SQLite's tests now run.  No impact to Chromium code.
 - 0002-Fix-expr.c-linker-error.patch is in upstream.
 - 0002-Use-seperate-page-cache-pools-for-each-sqlite-connec.patch had a minor
change.
 - 0005-Virtual-table-supporting-recovery-of-corrupted-datab.patch needed some
changes to the internal API used.  Change based on inspecting SQLite's code for
what it changed.
 - 0006-Custom-shell.c-helpers-to-load-Chromium-s-ICU-data.patch to fix a
linking issue.  I should remove this patch, nobody uses it.
 - 0010-Allow-read-only-memory-mapping.patch to follow SQLite's change in how
memory-mapped I/O works.  Was 0012.  I think this patch can be removed, I just
kept it to reduce free variables.
 - 0011-backport-Fix-collation-dequoting.patch was a backport, no longer needed.

AFAICT, it's all good.  I tried running through the various perf and memory
tests which failed last year, but found that even figuring out how to run many
of them was challenging enough that I punted.  Anything hard to run preemptively
just provides incentive to run it on demand instead.

[Scott Hess - ex-Googler, 2016-01-21 20:49:04]

On 2016/01/21 01:44:26, Scott Hess wrote:
> The trybots won't touch the amalgamation/sqlite3.c file, but I wrote a hack to
> split it up and ran that through the trybots here:
>    https://codereview.chromium.org/1560323002
> That may differ in minor ways from this, I'll hack it through again before
> landing.

OK, this CL's patch-set-1 is running through trybots at:
   https://codereview.chromium.org/1560323002/#ps40001
[Above CL's patch-set-5.]

[michaeln, 2016-01-21 21:01:02]

r/s lgtm !!!

[Scott Hess - ex-Googler, 2016-01-22 23:37:51]

On 2016/01/21 21:01:02, michaeln wrote:
> r/s lgtm !!!

Well, the try run looks green, except for some oilpan bots, but I can't think of
any way oilpan would negatively interact with this, and I've reached
exhaustion/not-caring on running tests locally, so gonna push it.  Thanks for
the lgtm...

[commit-bot: I haz the power, 2016-01-22 23:47:03]

Description was changed from

==========
Import SQLite 3.10.2.

Ran through the import script in third_party/sqlite/README.Chromium,
including the SQLite test suite.  No new errors observed.

SQLite changes are at http://www.sqlite.org/changes.html , Chromium
previously used 3.8.7.4.

All patches were applied and the results reviewed to make sure
backported patches were safe to remove, and retained patches were still
covering what was necessary.

Keep fts4 disabled, and also the new fts3 virtual table and unicode61
tokenizer.  Once enabled, these are very hard to disable, and there
doesn't seem to be any pressure to enable them.  Set
SQLITE_DEFAULT_PCACHE_INITSZ to 0 so SQLite does not preallocate 100
pages to pcache (both so Chromium can manage memory, also because our
code disables shared pcache).  Other SQLITE_* flags were reviewed for
applicability, none looked essential.

BUG=579743
==========

to

==========
Import SQLite 3.10.2.

Ran through the import script in third_party/sqlite/README.Chromium,
including the SQLite test suite.  No new errors observed.

SQLite changes are at http://www.sqlite.org/changes.html , Chromium
previously used 3.8.7.4.

All patches were applied and the results reviewed to make sure
backported patches were safe to remove, and retained patches were still
covering what was necessary.

Keep fts4 disabled, and also the new fts3 virtual table and unicode61
tokenizer.  Once enabled, these are very hard to disable, and there
doesn't seem to be any pressure to enable them.  Set
SQLITE_DEFAULT_PCACHE_INITSZ to 0 so SQLite does not preallocate 100
pages to pcache (both so Chromium can manage memory, also because our
code disables shared pcache).  Other SQLITE_* flags were reviewed for
applicability, none looked essential.

BUG=579743
R=michaeln@chromium.org

Committed: https://crrev.com/e0b6a82cadecbda7e005921c2416fc7afb3542aa
Cr-Commit-Position: refs/heads/master@{#371085}
==========

[commit-bot: I haz the power, 2016-01-22 23:47:06]

Patchset 1 (id:??) landed as
https://crrev.com/e0b6a82cadecbda7e005921c2416fc7afb3542aa
Cr-Commit-Position: refs/heads/master@{#371085}

[Scott Hess - ex-Googler, 2016-01-22 23:47:36]

Description was changed from

==========
Import SQLite 3.10.2.

Ran through the import script in third_party/sqlite/README.Chromium,
including the SQLite test suite.  No new errors observed.

SQLite changes are at http://www.sqlite.org/changes.html , Chromium
previously used 3.8.7.4.

All patches were applied and the results reviewed to make sure
backported patches were safe to remove, and retained patches were still
covering what was necessary.

Keep fts4 disabled, and also the new fts3 virtual table and unicode61
tokenizer.  Once enabled, these are very hard to disable, and there
doesn't seem to be any pressure to enable them.  Set
SQLITE_DEFAULT_PCACHE_INITSZ to 0 so SQLite does not preallocate 100
pages to pcache (both so Chromium can manage memory, also because our
code disables shared pcache).  Other SQLITE_* flags were reviewed for
applicability, none looked essential.

BUG=579743
R=michaeln@chromium.org

Committed: https://crrev.com/e0b6a82cadecbda7e005921c2416fc7afb3542aa
Cr-Commit-Position: refs/heads/master@{#371085}
==========

to

==========
Import SQLite 3.10.2.

Ran through the import script in third_party/sqlite/README.Chromium,
including the SQLite test suite.  No new errors observed.

SQLite changes are at http://www.sqlite.org/changes.html , Chromium
previously used 3.8.7.4.

All patches were applied and the results reviewed to make sure
backported patches were safe to remove, and retained patches were still
covering what was necessary.

Keep fts4 disabled, and also the new fts3 virtual table and unicode61
tokenizer.  Once enabled, these are very hard to disable, and there
doesn't seem to be any pressure to enable them.  Set
SQLITE_DEFAULT_PCACHE_INITSZ to 0 so SQLite does not preallocate 100
pages to pcache (both so Chromium can manage memory, also because our
code disables shared pcache).  Other SQLITE_* flags were reviewed for
applicability, none looked essential.

BUG=579743
R=michaeln@chromium.org

Committed:
https://chromium.googlesource.com/chromium/src/+/e0b6a82cadecbda7e005921c2416...
==========

[Scott Hess - ex-Googler, 2016-01-22 23:47:39]

Committed patchset #1 (id:1) manually as
e0b6a82cadecbda7e005921c2416fc7afb3542aa.