From: neil AT robots DOT ox DOT ac DOT uk (Neil Townsend) Newsgroups: comp.os.msdos.djgpp Subject: Re: GCC / RHIDE and CWSDMPI/CWSDPR0/Stubbing Date: Wed, 2 Sep 1998 13:28:41 GMT Organization: None evident here. Lines: 27 Message-ID: <1998Sep2.132841.28604@catorobots.ox.ac.uk> References: NNTP-Posting-Host: cato.robots.ox.ac.uk NNTP-Posting-Date: 2 Sep 1998 13:28:47 GMT Originator: neil AT cato To: djgpp AT delorie DOT com DJ-Gateway: from newsgroup comp.os.msdos.djgpp Precedence: bulk In comp.os.msdos.djgpp Eli Zaretskii writes: > >If you are asking for 32M plus some more in a single chunk, then this is >a well-known and documented behavior of the `malloc' implementation used >by DJGPP; it has nothing to do with the DPMI server. `malloc' rounds up >the size of the chunk you are asking for to the next integral power of 2, >so when you ask for 32MB + 1 byte, `malloc' actually asks the DPMI server >for 64MB, which you don't have. No, it's a lot of moderately sized blocks; I ask for more by asking for more blocks, no a single larger block; it fails in the process of allocating a collection of blocks. On failure, just after failing to allocate a block of either 240 or 480 bytes, _go32_dpmi_get_free_memory_information() reports that 16646144 bytes are still available (ie 16256K or 15.875M), so it's not exactly 16M left, but pretty close. Accoring to mem 107K of conventional and 2112K of extended memory is used of the total initial 48767K before we start; go32-v2 reports 46283K available for DPMI). Is there anything obvious I should try? Many thanks for all your answers so far, I hope all this is making sense and I'm not asking anything too stupid! Neil -- Neil Townsend +44 (1865) 273121 neil AT robots DOT ox DOT ac DOT uk