public inbox for gentoo-commits@lists.gentoo.org
 help / color / mirror / Atom feed
From: "Sergei Trofimovich" <slyfox@gentoo.org>
To: gentoo-commits@lists.gentoo.org
Subject: [gentoo-commits] repo/gentoo:master commit in: dev-libs/jemalloc/
Date: Sun, 21 Mar 2021 10:19:46 +0000 (UTC)	[thread overview]
Message-ID: <1616321970.f9f1a8c80671b4b1944226c84a5197183384a2ba.slyfox@gentoo> (raw)

commit:     f9f1a8c80671b4b1944226c84a5197183384a2ba
Author:     Sergei Trofimovich <slyfox <AT> gentoo <DOT> org>
AuthorDate: Sun Mar 21 10:19:30 2021 +0000
Commit:     Sergei Trofimovich <slyfox <AT> gentoo <DOT> org>
CommitDate: Sun Mar 21 10:19:30 2021 +0000
URL:        https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=f9f1a8c8

dev-libs/jemalloc: drop USE=hardened option

The only effect of USE=hardened was to pass a --disable-syscall
configure option. It's only effect is to use library wrappers
to memory allocation instead of direct syscalls to kernel.

As jemalloc overrides a bunch of library memory allocation
library calls itself this introduces extra chances to interact
badly with other tool that override library calls.

In #617518 sandbox became incompatible with jemalloc users and
produced deadlocks. Example from ruby[jemalloc]:

    malloc_init() [called by jemalloc]
     -> open("/proc/sys/vm/overcommit_memory")
      -> [sandbox takes over] open_DEFAULT()
       -> [sandbox initializes]
        -> init_env_entries ()
         -> _xmalloc()
          -> sb_mmap() [ try hard not to use olerloaded mmap() ]
           -> get_dlsym("mmap")
            -> calloc() [ sandbox loses control, overridden by jemalloc ]
             -> malloc_init() [ recursion, gets locked in attempt to
                  lock again locked malloc_init mutex ]

There should be no reason to use --disable-syscall on linux.
And especially not under generic (and enabled by default on
hardened profiles) USE=hardened.

The change drops --disable-syscall passing (and USE=hardened)
completely.

Users still can do it via EXTRA_ECONF= but there should be no
need for that. Upstream ./configure should do the right thing.

Reported-by: Maciej Piechotka
Closes: https://bugs.gentoo.org/617518
Package-Manager: Portage-3.0.17, Repoman-3.0.2
Signed-off-by: Sergei Trofimovich <slyfox <AT> gentoo.org>

 dev-libs/jemalloc/jemalloc-5.2.1-r1.ebuild | 50 ++++++++++++++++++++++++++++++
 1 file changed, 50 insertions(+)

diff --git a/dev-libs/jemalloc/jemalloc-5.2.1-r1.ebuild b/dev-libs/jemalloc/jemalloc-5.2.1-r1.ebuild
new file mode 100644
index 00000000000..e66e9eae517
--- /dev/null
+++ b/dev-libs/jemalloc/jemalloc-5.2.1-r1.ebuild
@@ -0,0 +1,50 @@
+# Copyright 1999-2021 Gentoo Authors
+# Distributed under the terms of the GNU General Public License v2
+
+EAPI="7"
+
+inherit autotools toolchain-funcs multilib-minimal
+
+DESCRIPTION="Jemalloc is a general-purpose scalable concurrent allocator"
+HOMEPAGE="http://jemalloc.net/ https://github.com/jemalloc/jemalloc"
+SRC_URI="https://github.com/jemalloc/jemalloc/releases/download/${PV}/${P}.tar.bz2"
+
+LICENSE="BSD"
+SLOT="0/2"
+KEYWORDS="~alpha ~amd64 ~arm ~arm64 ~hppa ~ia64 ~ppc ~ppc64 ~riscv ~s390 ~x86 ~amd64-linux ~x86-linux ~x64-macos ~x64-solaris"
+IUSE="debug lazy-lock prof static-libs stats xmalloc"
+HTML_DOCS=( doc/jemalloc.html )
+PATCHES=( "${FILESDIR}/${PN}-5.2.0-gentoo-fixups.patch" )
+
+MULTILIB_WRAPPED_HEADERS=( /usr/include/jemalloc/jemalloc.h )
+
+src_prepare() {
+	default
+	eautoreconf
+}
+
+multilib_src_configure() {
+	ECONF_SOURCE="${S}" \
+	econf  \
+		$(use_enable debug) \
+		$(use_enable lazy-lock) \
+		$(use_enable prof) \
+		$(use_enable stats) \
+		$(use_enable xmalloc)
+}
+
+multilib_src_install() {
+	# Copy man file which the Makefile looks for
+	cp "${S}/doc/jemalloc.3" "${BUILD_DIR}/doc" || die
+	emake DESTDIR="${D}" install
+}
+
+multilib_src_install_all() {
+	if [[ ${CHOST} == *-darwin* ]] ; then
+		# fixup install_name, #437362
+		install_name_tool \
+			-id "${EPREFIX}"/usr/$(get_libdir)/libjemalloc.2.dylib \
+			"${ED}"/usr/$(get_libdir)/libjemalloc.2.dylib || die
+	fi
+	use static-libs || find "${ED}" -name '*.a' -delete
+}


             reply	other threads:[~2021-03-21 10:19 UTC|newest]

Thread overview: 67+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-21 10:19 Sergei Trofimovich [this message]
  -- strict thread matches above, loose matches on Subject: below --
2024-02-25  0:01 [gentoo-commits] repo/gentoo:master commit in: dev-libs/jemalloc/ Joshua Kinard
2022-10-24  6:02 Joonas Niilola
2022-10-24  6:02 Joonas Niilola
2022-10-21 14:52 Arthur Zamarin
2022-10-21 14:52 Arthur Zamarin
2022-10-21 14:52 Arthur Zamarin
2022-10-21 14:52 Arthur Zamarin
2022-10-21 14:52 Arthur Zamarin
2022-10-21 12:02 Joonas Niilola
2022-10-21 12:02 Joonas Niilola
2022-09-26  7:31 Michał Górny
2022-09-19  6:42 Michał Górny
2022-08-30 16:05 Arthur Zamarin
2022-08-30  7:17 Arthur Zamarin
2022-07-31 11:08 Agostino Sarubbo
2022-07-30 14:31 Arthur Zamarin
2022-07-30 12:35 Arthur Zamarin
2022-07-30  9:51 Arthur Zamarin
2022-07-30  9:51 Arthur Zamarin
2022-07-30  7:05 Joonas Niilola
2022-07-30  7:05 Joonas Niilola
2022-07-03 17:09 James Le Cuirot
2022-07-02  7:02 Jakov Smolić
2022-07-01  6:16 Arthur Zamarin
2022-07-01  6:10 Arthur Zamarin
2022-07-01  6:10 Arthur Zamarin
2022-07-01  5:18 Arthur Zamarin
2022-06-30 22:12 Yixun Lan
2022-06-30 21:47 Sam James
2022-06-30  5:55 Joonas Niilola
2022-06-30  5:55 Joonas Niilola
2022-05-28 20:42 Jakov Smolić
2022-05-14  5:32 WANG Xuerui
2022-05-05  9:51 WANG Xuerui
2022-01-23  7:41 Joonas Niilola
2022-01-20 11:47 Sam James
2022-01-18 22:16 Sam James
2022-01-18 12:26 Sam James
2022-01-15 22:01 Sam James
2022-01-15 22:01 Sam James
2022-01-15 10:25 Jakov Smolić
2022-01-15 10:25 Jakov Smolić
2021-09-20 20:54 James Le Cuirot
2020-09-01  9:50 Yixun Lan
2019-12-08  4:50 Matt Turner
2019-11-21 17:17 Sergei Trofimovich
2019-11-14 11:55 Agostino Sarubbo
2019-11-13 16:07 Agostino Sarubbo
2019-11-13  8:43 Agostino Sarubbo
2019-11-13  8:16 Agostino Sarubbo
2019-11-13  7:48 Agostino Sarubbo
2019-11-13  7:39 Agostino Sarubbo
2019-11-12 17:47 Aaron Bauman
2019-08-07 20:26 Thomas Deutschmann
2019-08-07 20:26 Thomas Deutschmann
2018-12-12 12:57 Mikle Kolyada
2018-08-23  0:22 Thomas Deutschmann
2018-05-28  6:56 Jory Pratt
2018-04-09 15:01 Lars Wendler
2018-04-09 14:23 Lars Wendler
2017-12-03 17:50 Fabian Groffen
2017-08-29 14:14 Guilherme Amadio
2016-12-14  0:40 Jory Pratt
2016-07-17  8:15 Jory Pratt
2015-11-28 16:23 Jory Pratt
2015-11-26 21:58 Jory Pratt

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1616321970.f9f1a8c80671b4b1944226c84a5197183384a2ba.slyfox@gentoo \
    --to=slyfox@gentoo.org \
    --cc=gentoo-commits@lists.gentoo.org \
    --cc=gentoo-dev@lists.gentoo.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox