public inbox for cygwin-cvs@sourceware.org help / color / mirror / Atom feed
From: Corinna Vinschen <corinna@sourceware.org> To: cygwin-cvs@sourceware.org Subject: [newlib-cygwin] Cygwin: mmap_alloc: fix comment to document using the extended memory API Date: Tue, 7 Apr 2020 16:24:53 +0000 (GMT) [thread overview] Message-ID: <20200407162453.7A41A385DC1A@sourceware.org> (raw) https://sourceware.org/git/gitweb.cgi?p=newlib-cygwin.git;h=3fe9b02ccd6aeb684e39ca3fb9a6fd3ee5ede852 commit 3fe9b02ccd6aeb684e39ca3fb9a6fd3ee5ede852 Author: Corinna Vinschen <corinna@vinschen.de> Date: Tue Apr 7 14:17:04 2020 +0200 Cygwin: mmap_alloc: fix comment to document using the extended memory API Signed-off-by: Corinna Vinschen <corinna@vinschen.de> Diff: --- winsup/cygwin/mmap_alloc.cc | 23 +++++++++++++---------- 1 file changed, 13 insertions(+), 10 deletions(-) diff --git a/winsup/cygwin/mmap_alloc.cc b/winsup/cygwin/mmap_alloc.cc index b42bc16e1..cd4d49bc9 100644 --- a/winsup/cygwin/mmap_alloc.cc +++ b/winsup/cygwin/mmap_alloc.cc @@ -4,17 +4,20 @@ #include "mmap_alloc.h" #include <sys/param.h> -/* FIXME? Unfortunately the OS doesn't support a top down allocation with - a ceiling value. The ZeroBits mechanism only works for - NtMapViewOfSection and it only evaluates the high bit of ZeroBits - on 64 bit, so it's pretty much useless for our purposes. +/* Starting with Windows 10 1803 we use VirtualAlloc2 and MapViewOfFile3 + (or rather NtMapViewOfSectionEx), rather than the below class. + + Up to Windows 10 1709, the OS doesn't support a top down allocation with + a ceiling value. The ZeroBits mechanism only works for NtMapViewOfSection + and it only evaluates the high bit of ZeroBits on 64 bit, so it's pretty + much useless for our purposes. + + If the below simple mechanism to perform top-down allocations turns out to + be too dumb, we need something else. One idea is to divide the space in + (3835) 4 Gig chunks and just store the available free space per slot. Then + we can go top down from slot to slot and only try slots which are supposed + to have enough space. Bookkeeping would be very simple and fast. */ - If the below simple mechanism to perform top-down allocations - turns out to be too dumb, we need something else. One idea is to - dived the space in (3835) 4 Gig chunks and simply store the - available free space per slot. Then we can go top down from slot - to slot and only try slots which are supposed to have enough space. - Bookkeeping would be very simple and fast. */ PVOID mmap_allocator::alloc (PVOID in_addr, SIZE_T in_size, bool fixed) {
reply other threads:[~2020-04-07 16:24 UTC|newest] Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=20200407162453.7A41A385DC1A@sourceware.org \ --to=corinna@sourceware.org \ --cc=cygwin-cvs@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).