[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference bulova::decw_jan-89_to_nov-90

Title:	DECWINDOWS 26-JAN-89 to 29-NOV-90
Notice:	See 1639.0 for VMS V5.3 kit; 2043.0 for 5.4 IFT kit
Moderator:	STAR::VATNE

Created:	Mon Oct 30 1989
Last Modified:	Mon Dec 31 1990
Last Successful Update:	Fri Jun 06 1997
Number of topics:	3726
Total number of notes:	19516

973.0. "malloc -> XtFree memory mixup?" by ASD::DIGRAZIA () Mon Jun 19 1989 09:49

	We have code that allocates memory using malloc, and
	frees the same memory using XtFree.

	Is it ok to mix malloc and XtFree calls, or ought
	we use malloc/free or XtCalloc/XtFree consistently?

	The symptom we see is that after XtFree-ing some malloc'd
	memory, the next call to malloc returns a 0, but we know we 
	haven't run out of memory.

	Regards, Robert.

T.R	Title	User	Personal Name	Date	Lines
973.1		RAB::DESAI	Jatin Desai	`Mon Jun 19 1989 10:21`	9
	> Is it ok to mix malloc and XtFree calls, or ought > we use malloc/free or XtCalloc/XtFree consistently? I believe you can not mix-n-match malloc/XtMalloc and free/XtFree. You must use the right (matching) one. Jatin
973.2	While we're on the subject,	POOL::HALLYB	The Smart Money was on Goliath	`Mon Jun 19 1989 15:26`	8
	Can one use LIB$GET_VM on VMS systems as well as alloc/free ? Presuming, of course, the "free" routine corresponds to the "get" routine. Can be important when combining existing code into DW applications. John
973.3	Sure can	EVETPU::REINIG	This too shall change	`Mon Jun 19 1989 15:30`	5
	Sure can. TPU uses LIB$GET_VM. Since it calls the toolkit, it also uses alloc/free. FAKE_VM helps with all of this. August G. Reinig
973.4	difference?	BUFFER::RITTO		`Thu Jun 22 1989 10:12`	6
	What's the difference between using malloc and XtMalloc? Is there a reason to use one over the other? Confused, but thankful, Pat
973.5		CASEE::LACROIX	Gone with the wind	`Thu Jun 22 1989 10:42`	16
	> What's the difference between using malloc and XtMalloc? Is > there a reason to use one over the other? In general, XtMalloc/XtFree is supposed to be a lot faster than malloc/free in the C RTL. But the most important things is that some of the stuff you have XtMalloced maybe later XtFreed or XtRealloced by some toolkit code, or by some code you have stolen somewhere. XtFreeing something that was malloced is potentially a serious problem. The rules would be something like: when you are doing DECwindows programming, UI stuff, use XtMalloc/XtFree (or XMalloc/XFree...), and whenever you are interfacing with something else, THINK A LOT. It is very easy to shoot yourself in the foot. And yes, FAKE_VM is absolutely GREAT for memory corruptions (and you will have some). Denis.
973.6	Cheap plug for FAKE_VM	DDIF::BRAMHALL	Mark Bramhall, CDA architecture	`Tue Jun 27 1989 16:04`	7
	A cheap plug for FAKE_VM... FAKE_VM remembers the allocation method (LIB$GET_VM, malloc, XtMalloc, etc.) and will tell you if you are freeing the allocated memory with a different method. Freeing with a different method is one of the surest ways to totally corrupt your memory!
973.7	For us, I found trivial cause	GRANDE::DIGRAZIA		`Wed Jul 05 1989 11:30`	17
	Speaking of sure ways to corrupt memory, I entered the base note because we had semi-reproducible explosions in our code, which we decided were caused by something haywire in our dynamic-memory stuff. We knew we had a few malloc/XtFree combinations, etc. But it turns out that I found some simple, old-fashioned overruns in one of our modules. (I found a strcpy moving too-long strings into malloc'd memory, thereby overwriting malloc's bookkeeping words.) Many thanks for assistance. Regards, Robert.