home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!paladin.american.edu!darwin.sura.net!sgiblab!sgigate!sgi!fido!joker!myoung
- From: myoung@joker.asd.sgi.com (mark young)
- Newsgroups: comp.sys.sgi
- Subject: Re: GE FAULT:vertex buffer: 2 words
- Date: 23 Dec 1992 02:04:16 GMT
- Organization: Silicon Graphics, Inc. Mountain View, CA
- Lines: 73
- Distribution: usa
- Message-ID: <1h8hf0INN8ta@fido.asd.sgi.com>
- References: <72223@cup.portal.com>
- NNTP-Posting-Host: joker.asd.sgi.com
-
- In <72223@cup.portal.com> David_Anthony_Guevara@cup.portal.com writes:
- >I received the following error message in my SYSLOG
-
- > GE FAULT:vertex buffer: 2 words
- > 0001 0001
-
- >What is this error? Is it something I should be concerned with? As
- >far as I can tell, things are operating correctly. Any information
- >on what this is (or where I can find an explanation about it) would
- >be appreciated. Thanks!
-
- kurt akeley posted this quite a while ago. I don't think he'd mind if
- I reposted it (should find a place in the FAQ):
-
- > In article <1991May16.213625.17814@odin.corp.sgi.com>, scottm@thor.corp.sgi.com (Scott Machtmes) writes:
- > |> I have a customer that is getting the message " GE FAULT:vertex buffer:
- > |> 2 words. 0003 0000". Sometimes the message says 10 words. The message shows
- > |> up in the SYSLOG steadily. The GE4 and GM2 boards were replaced. At first the
- > |> graphics console was freezing up as well as giving the above message.
- > |>
- > |> Now the freezing has stopped but the messages are still occuring. Any
- > |> help would be appreciated. The system is a 240GTXB with 3.3.1 s/w.
- >
- > There was probably no need to replace the boards, these error messages are
- > almost certainly the result of driver error. On GT and GTX systems, when
- > a geometry engine detects an error condition, it reports a fault with a
- > two word sequence. The first word is the number of the geometry engine that
- > encountered the error condition. The second word is the error code.
- > Thus, in this case the error was detected by GE number 3, and has code
- > zero. A quick look in the GE3 source code indicates that error code
- > zero is generated when the vertex parser detects that the machine has
- > multiple branch codes set. A possible code sequence that would generate
- > this error condition is:
- >
- > bgnpolygon();
- > bgnline();
- > v3f(foo);
- >
- > For future reference, here's a list of all the error conditions that are
- > detected by GT and GTX geometry engines, and their meanings:
- >
- > GE number Fault number Meaning
- > --------- ------------ -------
- > 1 0 Out of memory, probably because too many matrixes
- > have been pushed onto the matrix stack (see below)
- > (User error)
- > 1 1 Matrix stack underflow (pop received with no
- > matrixes on the stack)
- > (User error)
- > 1,2,3,4,5 2 Bad command received
- > (System error)
- > 1,2,3,4,5 3 Internal microcode error (tried to return from
- > a subroutine when not in a subroutine)
- > (System error)
- > 3 0 Error detected in vertex mode. Probably caused
- > by calling multiple bgn* commands prior to a
- > vertex call.
- > (User error)
- > 4,5 4 RGB color commands received while in colorindex mode
- > (User error)
- > 4,5 5 Color index command received while in RGB mode
- > (User error)
- >
- > In the above (User error) means that the error probably results from an error
- > made by a GL programmer, and (System error) means that the error probably
- > results from a real bug, either hardware or system software. One can't be
- > sure, however.
- >
- > -- Kurt
- >
- > p.s. It would be great if this information was forwarded to the SEs in the
- > field, and kept in some place that can be referenced later!
- >
-