Date: Tue, 25 Nov 2008 23:26:23 +0100 From: Tijl Coosemans <tijl@ulyssis.org> To: freebsd-emulation@freebsd.org, Alex Kozlov <spam@rm-rf.kiev.ua> Cc: Jan Henrik Sylvester <me@janh.de>, Gerald Pfeifer <gerald@pfeifer.com> Subject: Re: wine-1.1.8 regression -- wine: could not load L"...": Invalid address Message-ID: <200811252326.25299.tijl@ulyssis.org> In-Reply-To: <alpine.LSU.1.99.0811252233340.3124@acrux.dbai.tuwien.ac.at> References: <20081124220311.GA14308@ravenloft.kiev.ua> <1227608432.1731.72.camel@localhost> <alpine.LSU.1.99.0811252233340.3124@acrux.dbai.tuwien.ac.at>
next in thread | previous in thread | raw e-mail | index | archive | help
Alex, On Tuesday 25 November 2008 22:39:25 Gerald Pfeifer wrote: > Alex, would you mind submitting this patch upstream as well, to=20 > wine-patches@winehq.org Cc:ing me on that? (If you prefer, I can > also do this, but since it's your patch...) >=20 > I'm currently travelling and will try to update our port in the > next couple of days; the earlier this gets upstream or at least > upstream review the better. It should probably first be discussed with Alexandre Julliard. The original code fails because it ends up allocating the virtual heap somewhere in the beginning of address space which probably interferes with other mappings. The patch moves this to (address_space_limit - 10 * VIRTUAL_HEAP_SIZE). I'm not sure that's correct. I think simply 0x80000000 would be better, but that's what Alexandre can tell you. Also, in the #else case you change a 1 into a 0. Is that intentional? =2D =A0 =A0if (!wine_mmap_enum_reserved_areas( alloc_virtual_heap, &heap_ba= se, 1 )) + =A0 =A0if (!wine_mmap_enum_reserved_areas( alloc_virtual_heap, &heap_base= , 0 ))
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200811252326.25299.tijl>