Re: [ql-developers] In search of a 68060RC 60MHz with 0E41J mask
On 24 Aug 2004 at 13:25, Richard Zidlicky wrote: > isn't RC the version without FPU and MMU? No, the RC is just a prefix to the speed grade. The 50 MHz version without FPU and MMU is 68EC060RC50 for example. MMU replaces EC by LC and MMU+FPU replaces EC or LC by nothing. Peter
Re: [ql-developers] In search of a 68060RC 60MHz with 0E41J mask
Thierry, I just read in the ql-users archive and saw that Fabrizio has the problem with crashing Qliberated programs, too. Since he has a 68LC060RC75, he must have a newer mask revision. I conclude that the CPU errata can hardly be the cause of problem you experience. Your dramatic public Q60 blame look even more inappropriate now. Peter
Re: [ql-developers] In search of a 68060RC 60MHz with 0E41J mask
On Tue, Aug 24, 2004 at 01:37:17AM +0200, Thierry Godefroy wrote: > MC68060RC60 are available... MC, not XC. Of course, I can't prove they > were available back in 2001, but that's beyond the point as even a RC50 > could have been overclocked to 66MHz... isn't RC the version without FPU and MMU? > I don't care about other processors, but if you remember all the fuss that > was made about just -one- bug in the Pentium FPU (the "divide bug", I bet > that Motorola was lucky that their processor was not as popular as the > Pentium... that would seem so. It is not just the 68060, 68040 has its problems too. As soon as you do things like MMU and cache your path is paved with workarounds for CPU bugs - and those guys who were doing SMP with m68k machines must really have white hair by now. Not to mention that the basic design of the MMU and cache is already extremely unforgiving.. Linux on 68060 has been debugged for almost a decade now and we are still finding cache and MMU related bugs. > I'm searching the proofs, but when you got a processor that can choke on as > simple a code as: > >.../... >Bcc LABEL_BANG >.../... > > LABEL_BANG MOVE An,USP <--- I11 errata: USP corrupted ! its SV mode code so you can reasonably work around by software. Have you actually seen QDOS code that would trigger that? The really worst bug is I16, except that is likely to be triggered only by intent. I did repeatedly look through the other bugs and don't think they would be triggered in QDOS code. Eg I could not think of any real life conditions where QDOS code would trigger I14.. now gcc has advanced to stage of producing such broken code. Richard
Re: [ql-developers] In search of a 68060RC 60MHz with 0E41J mask
Thierry, > > Probably all full-blown 66 MHz Q60 will have that mask revision. There is > > nothing wrong about that. Especially as you have the "A" labelled > > 68060RC60, which should be the best full-blown 68060 silicon ever for > > overclocking. > > I don't know were you got that info (or impression) from, but the XC68060 > is said by Amiga folks to be _less_ overclockable than the MC68060. Example > : http://www.amiga-hardware.com/mc68060.html Citation: "The MC versions > tends to be much more tolerant of heat and is >more tolerant of being overclocked." Known. Goes among RC50 speed grade, but you were referring to RC60. > I also browsed many Amiga sites today, and all the folks there use 50MHz > MC68060s to overclock them at 66MHz without a single problem... Also known. I tried to supply you with optimal savety margin. Wasted effort, as it looks now. But I wonder what had happened, if I provided you a slower speed grade with later mask revision. Probably a public lamentation about the speed grade :-) > > (The only full-blown chips leaving Motorola fabs at the time I built your > > board were 50 MHz, and I'm quite sure that was not what you wanted.) > > http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MC68060&node > Id=018rH3YTLC4622#orderables At the time I built your board, it was definitely _not_ on the website. I remember that exactly, because I wondered about it. Also I'd be careful to conclude availability from that page. Try to actually order and see what happens. > The 1G65V mask _was_ normal mainstream production... It has been replaced > in 1999 with the 0E41J, at which point the processor was renamed from > XC68060 into MC68060: XC is used by Motorola for "Pilot production device" > while MC is used for "Fully qualified parts", clearly indicating that prior > to 0E41J maskset, the 68060 was still considered "experimental" by > Motorola. Same ole, same ole. Motorola often leaves chips labelled XC for long, while producing them fullscale. Some never get labelled MC. By itself, that means nothing. > I don't know what 'FUD' means, but I'm simply searching for a bug-less > 68060. FUD = "Fear Uncertainty Doubt". You were not simply searching, but made a dramatic exaggeration out of this, just creating inappropiate feelings of uncertainty among Q60 users. > I don't care about other processors, but if you remember all the fuss that > was made about just -one- bug in the Pentium FPU (the "divide bug", I bet > that Motorola was lucky that their processor was not as popular as the > Pentium... On the other hand, the MC68060, which is the "fully qualified" > device, is bugless, and the buggy XC68060 was clearly just a pre-release... Motorola has sold the majority of 68060 this way, I see this as mainstream production. > One thing is certain: I will not loose more time > trying to make the copyback mode work under SMSQ/E, and this until my Q60 > is fitted with a bug-less processor: perhaps will the bug prove to be > purely a software one in the end, You could have mentioned that instead of just making a big fuss against my hardware. > but my free time is too precious to loose > it trying to distinguish what is the processor's fault and the software's > one, and review the machine code of hundreds of kilobytes of OS extensions, > toolkits and the like to spot the (correct !) line which fails to execute > on my buggy processor... If you had actually placed your promised order for a Q60/80 in time, even that scenario (which I still find pretty unrealistic) wouldn't exist. Also you could have asked me, but I gather that you prefer to stir up the public first. > > I doubt such chips are on the market. > > Again: > > http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MC68060&node > Id=018rH3YTLC4622#orderables Again: See above. > I my case the only > "catastrophic accident" is an added expense in order to end up with a 100% > working system. I can deal with it. You should deal with your mistake too, > like I do with mine. No need to whine. I don't. I clearly reject your claim not to have a fully working system. "Working" needs to be defined by the normal purpose of a good, which is 100% fulfilled by your Q60. Your hypothetical added expense (please let me know if a RC60 with the mask revision you want exists on the market at all) would just provide you an extravagance that other mainstream 68060 systems have not. Chip documentation (including errata) has to be accepted if you're a specialist dealing with lowest level software. That's about all. Peter
Re: [ql-developers] In search of a 68060RC 60MHz with 0E41J mask
On Mon, 23 Aug 2004 23:45:18 +0200, Peter Graf wrote: > On 23 Aug 2004 at 15:22, Thierry Godefroy wrote: > > > I'm quite upset !... > > > > I just discovered that my Q60 was fitted with a buggy 68060 ! > > Yes, a pre-1996 mask 060 (1G65V mask), while my Q60 was built > > as bought in 2001 !!!... This makes me wonder how many among > > the Q60s have that same buggy processor fitted... > > Probably all full-blown 66 MHz Q60 will have that mask revision. There is > nothing wrong about that. Especially as you have the "A" labelled > 68060RC60, which should be the best full-blown 68060 silicon ever for > overclocking. I don't know were you got that info (or impression) from, but the XC68060 is said by Amiga folks to be _less_ overclockable than the MC68060. Example : http://www.amiga-hardware.com/mc68060.html Citation: "The MC versions tends to be much more tolerant of heat and is more tolerant of being overclocked." I also browsed many Amiga sites today, and all the folks there use 50MHz MC68060s to overclock them at 66MHz without a single problem... > I toiled hard to locate them, and you could be a proud owner > instead of lamenting. I didn't accuse -you- and I didn't lament either. I said I was (and still am) upset. I'm upset, because when my board was built, full blown and -bugless- MC68060 existed and I end up with a buggy processor that I will have to replace. I'm upset because Motorola continued to sell buggy processors from their stock. I'm upset because I was stupid enough to believe that I would not have to worry about such bugs. I'm uspet because I didn't check in the available docs before ordering the card and as a result could not check with you what processor was to be used. > (The only full-blown chips leaving Motorola fabs at the time I built > your board were 50 MHz, and I'm quite sure that was not what you wanted.) http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MC68060&nodeId=018rH3YTLC4622#orderables MC68060RC60 are available... MC, not XC. Of course, I can't prove they were available back in 2001, but that's beyond the point as even a RC50 could have been overclocked to 66MHz... > The 1G65V mask revision is normal mainstream production, The 1G65V mask _was_ normal mainstream production... It has been replaced in 1999 with the 0E41J, at which point the processor was renamed from XC68060 into MC68060: XC is used by Motorola for "Pilot production device" while MC is used for "Fully qualified parts", clearly indicating that prior to 0E41J maskset, the 68060 was still considered "experimental" by Motorola. References: Product identification: http://www.cpu-world.com/info/id/Motorola-identification.html 0E41J product change notice: http://www.freescale.com/files/shared/doc/pcn/945048375992collateral.html > the 50 MHz speed > grade is used in many thousands of Amiga and VME machines. The errata are > business as usual, no need to spread public FUD here. I don't know what 'FUD' means, but I'm simply searching for a bug-less 68060. I'm not trying to spread anything neither to badmouth you. My message was simply here to both warn people that they could be in the same situation as mine and to ask for anyone with an available replacement processor. At least, my warning will have the positive effect to get the future versions of SMSQ/E fixed for the I14/I15 errata... > > I discovered this because Linux v2.4 got a recent fix for the > > I14 errata of these buggy processor, and logs the activation > > of the fix when the kernel boots. This fix consists in setting > > the bit 5 in the PCR (Processor Control Register), which > > disables one of the optimisations of the 060. But there are no > > less than 21 bugs in the maskset of my 060 !!!... > > If you're already upset about 21 bugs, I wonder how you'd deal with modern > chips that have far more bugs. Like a PowerPC derivative to which I've been > porting eCos recently :-( I don't care about other processors, but if you remember all the fuss that was made about just -one- bug in the Pentium FPU (the "divide bug", I bet that Motorola was lucky that their processor was not as popular as the Pentium... On the other hand, the MC68060, which is the "fully qualified" device, is bugless, and the buggy XC68060 was clearly just a pre-release... > > No wonder why > > I get weird results each time I try to modify SMSQ/E and QLib_run > > in order make them work properly in copyback mode (currently, > > QLiberated programs hang/crash/loop randomly when copyback is > > enabled). > > Any proof or just guessing? Reminds me of another "hardware bug" which was > lamented in public, but turned out a MMU initialization error in SMSQ/E > which caused random effects. I'm searching the proofs, but when you got a processor that can choke on as simple a code as: .../... Bcc LABEL_BANG .../... LABEL_BANG MOVE An,USP <--- I11 errata: USP corrupted !
Re: [ql-developers] In search of a 68060RC 60MHz with 0E41J mask
On Mon, Aug 23, 2004 at 03:22:24PM +0200, Thierry Godefroy wrote: > > I'm quite upset !... > > I just discovered that my Q60 was fitted with a buggy 68060 ! > Yes, a pre-1996 mask 060 (1G65V mask), while my Q60 was built > as bought in 2001 !!!... This makes me wonder how many among > the Q60s have that same buggy processor fitted... > > I discovered this because Linux v2.4 got a recent fix for the > I14 errata of these buggy processor, and logs the activation > of the fix when the kernel boots. didn't look at it yet, does Linux actually test whether its an affected CPU, or just disable that optimisation? > This fix consists in setting > the bit 5 in the PCR (Processor Control Register), which > disables one of the optimisations of the 060. But there are no > less than 21 bugs in the maskset of my 060 !!!... No wonder why > I get weird results each time I try to modify SMSQ/E and QLib_run > in order make them work properly in copyback mode (currently, > QLiberated programs hang/crash/loop randomly when copyback is > enabled). I did once go through all the horrible 68060 errata and it doesn't appear SMSQ or QDOS would trigger any of these. > For the ones interested in the details about these bugs, they > will find the 68060 device errata document here: > http://www.freescale.com/files/32bit/doc/no_sub_type/MC68060DE.pdf > > Anyway... > > I'm now in search for a bug-less and full blown 60MHz 68060 > (0E41J mask ONLY)... does that exist? I think it exists only on paper or as RC/LC version :(( If you look at 68040 and 68060 errata lists you would be surprised that those CPUs ever worked, but apparently all but one of those bugs can be "managed" by OS software without significant performance degradation. Richard
Re: [ql-developers] In search of a 68060RC 60MHz with 0E41J mask
On 23 Aug 2004 at 15:22, Thierry Godefroy wrote: > I'm quite upset !... > > I just discovered that my Q60 was fitted with a buggy 68060 ! > Yes, a pre-1996 mask 060 (1G65V mask), while my Q60 was built > as bought in 2001 !!!... This makes me wonder how many among > the Q60s have that same buggy processor fitted... Probably all full-blown 66 MHz Q60 will have that mask revision. There is nothing wrong about that. Especially as you have the "A" labelled 68060RC60, which should be the best full-blown 68060 silicon ever for overclocking. I toiled hard to locate them, and you could be a proud owner instead of lamenting. (The only full-blown chips leaving Motorola fabs at the time I built your board were 50 MHz, and I'm quite sure that was not what you wanted.) The 1G65V mask revision is normal mainstream production, the 50 MHz speed grade is used in many thousands of Amiga and VME machines. The errata are business as usual, no need to spread public FUD here. > I discovered this because Linux v2.4 got a recent fix for the > I14 errata of these buggy processor, and logs the activation > of the fix when the kernel boots. This fix consists in setting > the bit 5 in the PCR (Processor Control Register), which > disables one of the optimisations of the 060. But there are no > less than 21 bugs in the maskset of my 060 !!!... If you're already upset about 21 bugs, I wonder how you'd deal with modern chips that have far more bugs. Like a PowerPC derivative to which I've been porting eCos recently :-( > No wonder why > I get weird results each time I try to modify SMSQ/E and QLib_run > in order make them work properly in copyback mode (currently, > QLiberated programs hang/crash/loop randomly when copyback is > enabled). Any proof or just guessing? Reminds me of another "hardware bug" which was lamented in public, but turned out a MMU initialization error in SMSQ/E which caused random effects. > I'm now in search for a bug-less and full blown 60MHz 68060 > (0E41J mask ONLY)... I doubt such chips are on the market. Thanks for all the flowers, maybe build your own error-free hardware next time. Or choose one of the million selling companies which will surely serve you better than me. Peter