public inbox for glibc-bugs@sourceware.org help / color / mirror / Atom feed
From: "sstewartgallus00 at mylangara dot bc.ca" <sourceware-bugzilla@sourceware.org> To: glibc-bugs@sourceware.org Subject: [Bug nptl/17214] Expose a clone variant that shares stacks instead of jumping to a new one Date: Mon, 26 Jan 2015 22:33:00 -0000 [thread overview] Message-ID: <bug-17214-131-ZxFzAJABnf@http.sourceware.org/bugzilla/> (raw) In-Reply-To: <bug-17214-131@http.sourceware.org/bugzilla/> https://sourceware.org/bugzilla/show_bug.cgi?id=17214 --- Comment #15 from Steven Stewart-Gallus <sstewartgallus00 at mylangara dot bc.ca> --- It occurs to me that maybe clone with vfork could be exposed in a safe manner along the lines of the following (although GLibc would probably implement it in assembly directly). __attribute__((noinline)) __attribute__((noclone)) __attribute__((no_sanitize_address)) static pid_t safe_vfork( int (*volatile f)(void *), void *volatile arg) { __atomic_signal_fence(__ATOMIC_SEQ_CST); pid_t child = vfork(); if (0 == child) _Exit(f(arg)); return child; } Aside from the weirdness of vfork this should be no less safer than something along the lines of the following: __attribute__((noinline)) __attribute__((noclone)) __attribute__((no_sanitize_address)) static pid_t safe_vclone( int volatile clone_flags, int (*volatile f)(void *), void *volatile arg) { long maybe_page_size = sysconf(_SC_PAGE_SIZE); assert(maybe_page_size >= 0); long maybe_stack_min_size = sysconf(_SC_THREAD_STACK_MIN); assert(maybe_stack_min_size >= 0); size_t page_size = maybe_page_size; size_t stack_min_size = maybe_stack_min_size; /* We need an extra page for signals */ size_t stack_size = stack_min_size + page_size; size_t stack_and_guard_size = page_size + stack_size + page_size; void *child_stack = mmap( 0, stack_and_guard_size, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANONYMOUS | MAP_GROWSDOWN | MAP_STACK, -1, 0); if (0 == child_stack) return -1; /* Guard pages are shared between the stacks */ if (-1 == mprotect((char *)child_stack, page_size, PROT_NONE)) goto on_err; if (-1 == mprotect((char *)child_stack + page_size + stack_size, page_size, PROT_NONE)) goto on_err; void *stack_start = (char *)child_stack + page_size + stack_size; __atomic_signal_fence(__ATOMIC_SEQ_CST); pid_t child = clone(f, stack_start, clone_flags | CLONE_VM | CLONE_VFORK, arg); if (-1 == child) goto on_err; munmap(child_stack, stack_and_guard_size); return child; on_err: ; int errnum = errno; munmap(child_stack, stack_and_guard_size); errno = errnum; return -1; } -- You are receiving this mail because: You are on the CC list for the bug.
prev parent reply other threads:[~2015-01-26 22:33 UTC|newest] Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top 2014-07-29 21:23 [Bug nptl/17214] New: Expose a function to reset the PID cache sstewartgallus00 at mylangara dot bc.ca 2014-07-30 4:07 ` [Bug nptl/17214] " carlos at redhat dot com 2014-07-30 17:39 ` sstewartgallus00 at mylangara dot bc.ca 2014-07-30 17:41 ` sstewartgallus00 at mylangara dot bc.ca 2014-07-31 23:25 ` carlos at redhat dot com 2014-08-01 21:59 ` sstewartgallus00 at mylangara dot bc.ca 2014-08-26 4:38 ` bugdal at aerifal dot cx 2014-08-26 18:11 ` sstewartgallus00 at mylangara dot bc.ca 2014-08-26 18:31 ` sstewartgallus00 at mylangara dot bc.ca 2014-10-30 22:01 ` rickyz at chromium dot org 2014-12-18 0:15 ` jld at mozilla dot com 2014-12-18 23:58 ` [Bug nptl/17214] Expose a clone variant that shares stacks instead of jumping to a new one sstewartgallus00 at mylangara dot bc.ca 2014-12-19 0:13 ` sstewartgallus00 at mylangara dot bc.ca 2014-12-19 1:17 ` bugdal at aerifal dot cx 2014-12-19 20:29 ` sstewartgallus00 at mylangara dot bc.ca 2014-12-19 21:01 ` bugdal at aerifal dot cx 2015-01-26 22:33 ` sstewartgallus00 at mylangara dot bc.ca [this message]
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=bug-17214-131-ZxFzAJABnf@http.sourceware.org/bugzilla/ \ --to=sourceware-bugzilla@sourceware.org \ --cc=glibc-bugs@sourceware.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: linkBe 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; as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).