summaryrefslogtreecommitdiff
blob: b4c5be8aa0cfbe318ba59a66cc691be9938d3152 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
# Copyright 1999-2010 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2
# $Header: /var/cvsroot/gentoo-x86/dev-db/mysql/mysql-5.1.44-r3.ebuild,v 1.3 2010/04/01 20:41:21 robbat2 Exp $

MY_EXTRAS_VER="20100324-0213Z"
EAPI=2

# PBXT
PBXT_VERSION='1.0.10-rc'
# XtraDB
# handler/i_s.cc: In function 'int i_s_innodb_buffer_pool_pages_index_fill(THD*, TABLE_LIST*, COND*)':
# handler/i_s.cc:801: error: invalid conversion from 'const char*' to 'char*'
# handler/i_s.cc: In function 'int i_s_innodb_admin_command_fill(THD*, TABLE_LIST*, COND*)':
# handler/i_s.cc:2995: warning: deprecated conversion from string constant to 'char*'
# PERCONA_VER='5.1.42-9' XTRADB_VER='1.0.6-9'

inherit toolchain-funcs mysql
# only to make repoman happy. it is really set in the eclass
IUSE="$IUSE"

# REMEMBER: also update eclass/mysql*.eclass before committing!
KEYWORDS="~alpha ~amd64 ~arm ~hppa ~ia64 ~ppc ~ppc64 ~s390 ~sh ~sparc ~x86 ~sparc-fbsd ~x86-fbsd"

# When MY_EXTRAS is bumped, the index should be revised to exclude these.
EPATCH_EXCLUDE=''

DEPEND="|| ( >=sys-devel/gcc-3.4.6 >=sys-devel/gcc-apple-4.0 )"
RDEPEND="!media-sound/amarok[embedded]"

# Please do not add a naive src_unpack to this ebuild
# If you want to add a single patch, copy the ebuild to an overlay
# and create your own mysql-extras tarball, looking at 000_index.txt

# Official test instructions:
# USE='berkdb -cluster embedded extraengine perl ssl community' \
# FEATURES='test userpriv -usersandbox' \
# ebuild mysql-X.X.XX.ebuild \
# digest clean package
src_test() {
	# Bug #213475 - MySQL _will_ object strenously if your machine is named
	# localhost. Also causes weird failures.
	[[ "${HOSTNAME}" == "localhost" ]] && die "Your machine must NOT be named localhost"

	emake check || die "make check failed"
	if ! use "minimal" ; then
		if [[ $UID -eq 0 ]]; then
			die "Testing with FEATURES=-userpriv is no longer supported by upstream. Tests MUST be run as non-root."
		fi
		has usersandbox $FEATURES && eerror "Some tests may fail with FEATURES=usersandbox"
		cd "${S}"
		einfo ">>> Test phase [test]: ${CATEGORY}/${PF}"
		local retstatus_unit
		local retstatus_ns
		local retstatus_ps
		local t
		addpredict /this-dir-does-not-exist/t9.MYI

		# Ensure that parallel runs don't die
		export MTR_BUILD_THREAD="$((${RANDOM} % 100))"

		# archive_gis really sucks a lot, but it's only relevant for the
		# USE=extraengines case
		case ${PV} in
			5.0.42)
			mysql_disable_test "archive_gis" "Totally broken in 5.0.42"
			;;

			5.0.4[3-9]|5.0.[56]*|5.0.70|5.0.87)
			[ "$(tc-endian)" == "big" ] && \
			mysql_disable_test \
				"archive_gis" \
				"Broken in 5.0.43-70 and 5.0.87 on big-endian boxes only"
			;;
		esac

		# This was a slight testcase breakage when the read_only security issue
		# was fixed.
		case ${PV} in
			5.0.54|5.0.51*)
			mysql_disable_test \
				"read_only" \
				"Broken in 5.0.51-54, output in wrong order"
			;;
		esac

		# Ditto to read_only
		[ "${PV}" == "5.0.51a" ] && \
			mysql_disable_test \
				"view" \
				"Broken in 5.0.51, output in wrong order"

		# x86-specific, OOM issue with some subselects on low memory servers
		[ "${PV}" == "5.0.54" ] && \
			[ "${ARCH/x86}" != "${ARCH}" ] && \
			mysql_disable_test \
				"subselect" \
				"Testcase needs tuning on x86 for oom condition"

		# Broke with the YaSSL security issue that didn't affect Gentoo.
		[ "${PV}" == "5.0.56" ] && \
			for t in openssl_1 rpl_openssl rpl_ssl ssl \
				ssl_8k_key ssl_compress ssl_connect ; do \
				mysql_disable_test \
					"$t" \
					"OpenSSL tests broken on 5.0.56"
			done

		# New test was broken in first time
		# Upstream bug 41066
		# http://bugs.mysql.com/bug.php?id=41066
		[ "${PV}" == "5.0.72" ] && \
			mysql_disable_test \
				"status2" \
				"Broken in 5.0.72, new test is broken, upstream bug #41066"

		# The entire 5.0 series has pre-generated SSL certificates, they have
		# mostly expired now. ${S}/mysql-tests/std-data/*.pem
		# The certs really SHOULD be generated for the tests, so that they are
		# not expiring like this. We cannot do so ourselves as the tests look
		# closely as the cert path data, and we do not have the CA key to regen
		# ourselves. Alternatively, upstream should generate them with at least
		# 50-year validity.
		#
		# Known expiry points:
		# 4.1.*, 5.0.0-5.0.22, 5.1.7: Expires 2013/09/09
		# 5.0.23-5.0.77, 5.1.7-5.1.22?: Expires 2009/01/27
		# 5.0.78-5.0.90, 5.1.??-5.1.42: Expires 2010/01/28
		#
		# mysql-test/std_data/untrusted-cacert.pem is MEANT to be
		# expired/invalid.
		case ${PV} in
			5.0.*|5.1.*|5.4.*|5.5.*)
				for t in openssl_1 rpl_openssl rpl.rpl_ssl rpl.rpl_ssl1 ssl ssl_8k_key \
					ssl_compress ssl_connect rpl.rpl_heartbeat_ssl ; do \
					mysql_disable_test \
						"$t" \
						"These OpenSSL tests break due to expired certificates"
				done
			;;
		esac

		# These are also failing in MySQL 5.1 for now, and are believed to be
		# false positives:
		#
		# main.mysql_comment, main.mysql_upgrade, main.information_schema,
		# funcs_1.is_columns_mysql funcs_1.is_tables_mysql funcs_1.is_triggers:
		# fails due to USE=-latin1 / utf8 default
		#
		# main.mysql_client_test:
		# segfaults at random under Portage only, suspect resource limits.
		#
		# main.not_partition:
		# Failure reason unknown at this time, must resolve before package.mask
		# removal FIXME
		case ${PV} in
			5.1.*|5.4.*|5.5.*)
			for t in main.mysql_client_test main.mysql_comments \
				main.mysql_upgrade  \
				main.information_schema \
				main.not_partition funcs_1.is_columns_mysql \
				funcs_1.is_tables_mysql funcs_1.is_triggers; do
				mysql_disable_test  "$t" "False positives in Gentoo"
			done
			;;
		esac

		use profiling && use community \
		|| mysql_disable_test main.profiling \
			"Profiling test needs profiling support"

		if [ "${PN}" == "mariadb" ]; then
			for t in \
				parts.part_supported_sql_func_ndb \
				parts.partition_auto_increment_ndb ; do
					mysql_disable_test $t "ndb not supported in mariadb"
			done
		fi

		# create directories because mysqladmin might make out of order
		mkdir -p "${S}"/mysql-test/var-{ps,ns}{,/log}

		# We run the test protocols seperately
		emake test-unit
		retstatus_unit=$?
		[[ $retstatus_unit -eq 0 ]] || eerror "test-unit failed"

		emake test-ns force="--force --vardir=${S}/mysql-test/var-ns"
		retstatus_ns=$?
		[[ $retstatus_ns -eq 0 ]] || eerror "test-ns failed"
		has usersandbox $FEATURES && eerror "Some tests may fail with FEATURES=usersandbox"

		emake test-ps force="--force --vardir=${S}/mysql-test/var-ps"
		retstatus_ps=$?
		[[ $retstatus_ps -eq 0 ]] || eerror "test-ps failed"
		has usersandbox $FEATURES && eerror "Some tests may fail with FEATURES=usersandbox"

		# TODO:
		# When upstream enables the pr and nr testsuites, we need those as well.

		# Cleanup is important for these testcases.
		pkill -9 -f "${S}/ndb" 2>/dev/null
		pkill -9 -f "${S}/sql" 2>/dev/null
		failures=""
		[[ $retstatus_unit -eq 0 ]] || failures="${failures} test-unit"
		[[ $retstatus_ns -eq 0 ]] || failures="${failures} test-ns"
		[[ $retstatus_ps -eq 0 ]] || failures="${failures} test-ps"
		has usersandbox $FEATURES && eerror "Some tests may fail with FEATURES=usersandbox"
		[[ -z "$failures" ]] || die "Test failures: $failures"
		einfo "Tests successfully completed"
	else
		einfo "Skipping server tests due to minimal build."
	fi
}