<rqou>
oh, so you already need asbestos abatement anyways
<azonenberg_work>
welll
<azonenberg_work>
I thought that's all there was (the rest of the ceilings upstairs had traces but below regulated limits, the downstairs was undetectable
<rqou>
i thought "below regulated limits" was still bad?
<azonenberg_work>
It is, but it means it's not a hazmat if i chuck it myself :p
<rqou>
ah lol
<rqou>
i'm not even sure my normal respirator is suitable or not
<azonenberg_work>
I'm wearing a tyvek bunny suit and P100 respirator for sheetrock demo in general so i'm not super concerned about a few traces
<azonenberg_work>
My original plan was to RFQ for abatement of the closets, but most likely just mount a second layer of drywall over them and encapsulate them
<azonenberg_work>
i'd have to disclose if/when i sold the place but it would be sealed in and unable to leak fibers or be accidentally damaged
<rqou>
btw is a p100 filter actually good enough?
<azonenberg_work>
Yes
<azonenberg_work>
I'd want to clean-shave to get a good fit though
<rqou>
just be asian and there won't be a problem anymore :P
<azonenberg_work>
I shortened my beard to make the respirator fit better when doing sheetrock demo but didnt bother with a super close shave
<azonenberg_work>
because sheetrock dust is nuisance dust, not full hazmat
<azonenberg_work>
So if a little got in, whatever
<azonenberg_work>
anyway, the ceiling downstairs is especially problematic because of the placement right next to the main house hvac intake
<azonenberg_work>
So i'm definitely going to want to get that professionally handled
<azonenberg_work>
its only about 20 square feet, it shouldnt be too pricey
<azonenberg_work>
The closets are each about that big and there's something like 3 or 4 of them
<azonenberg_work>
Then the rest of the upstairs is mostly negative but there's a few bumpy spots that tested positive whewre the previous abatement guy was sloppy (guessing it was a DIY job decades ago)
<azonenberg_work>
(but below regulated limits)
<azonenberg_work>
The only thing i'd consider DIY removal of myself is that
<azonenberg_work>
Basically just pull the whole piece of sheetrock down and avoid pulverizing that area
<azonenberg_work>
then hepa vac everything to get any residue
<rqou>
hmm, wikipedia[citation needed] claims that a hepa vac isn't sufficient
azonenberg_work has quit [Ping timeout: 240 seconds]
Ishan_Bansal has quit [*.net *.split]
reportingsjr has quit [*.net *.split]
sielicki has quit [*.net *.split]
lain has quit [*.net *.split]
cyrozap has quit [*.net *.split]
finsternis has quit [*.net *.split]
ovf has quit [*.net *.split]
kuldeep has quit [*.net *.split]
cyrozap has joined ##openfpga
reportingsjr has joined ##openfpga
kuldeep has joined ##openfpga
finsternis has joined ##openfpga
lain has joined ##openfpga
Ishan_Bansal has joined ##openfpga
ovf has joined ##openfpga
sielicki has joined ##openfpga
azonenberg_work has joined ##openfpga
<ZipCPU>
rqou: Thanks for the reddit post!
implr has joined ##openfpga
digshadow has quit [Ping timeout: 264 seconds]
jhol has quit [*.net *.split]
awygle has quit [*.net *.split]
marex-cloud has quit [*.net *.split]
nickjohnson has quit [*.net *.split]
nurelin has quit [*.net *.split]
jhol has joined ##openfpga
nurelin has joined ##openfpga
nickjohnson has joined ##openfpga
marex-cloud has joined ##openfpga
awygle has joined ##openfpga
digshadow has joined ##openfpga
jhol has quit [Quit: Coyote finally caught me]
jhol has joined ##openfpga
Bike has quit [Quit: Lost terminal]
<rqou>
wow, amazon's recommendations are totally broken for me
<rqou>
it keeps thinking i want to buy bags of sulfur
<rqou>
whitequark is this your fault? :P
<qu1j0t3>
the feds are trolling you
<qu1j0t3>
don't take the bait
<rqou>
sulfur isn't illegal?
<qu1j0t3>
it's a gateway drug
<sorear>
sulfur does not seem to be on the list, but if you burn it and dissolve the vapors in water, that _will_ produce a controlled substance because wtf
<azonenberg>
i've had this happen before both here and in other chans
<azonenberg>
and i dont have a good solution
<azonenberg>
it might also be that whatever freenode server the bot is trying to talk to is down / ddos'd / netsplit
<azonenberg>
sorear: it's not controlled like you think when you hear of controlled narcotics
<azonenberg>
it's just that they keep an eye on who buys large volumes and purchases above some threshold (which is huge) have to be reported
<azonenberg>
The DEA isnt making lists of everyone who buys acetone :p
<azonenberg>
i mean maybe they are, but they dont do much with that info b/c the list is so long
<sorear>
NSA probably is though :p
<rqou>
home depot actually "checked id" last time i went to buy acetone
<azonenberg>
:o
<rqou>
which was kinda surprising to me
<azonenberg>
Honestly
<azonenberg>
i think they'd be more concerned about kids making TATP
<azonenberg>
than drugs
<azonenberg>
:p
<rqou>
by "checked id" i meant "the person staffing the self-checkout glanced at me, saw that i obviously wasn't <18 anymore, and didn't actually check my id"
<azonenberg>
Yeah exactly
<rqou>
yeah, i assume it's because of TATP
<azonenberg>
having a 18+ restriction is sensible on easily misused / hazardous things
<azonenberg>
making a full log of purchases is a bit more absurd
<rqou>
unless you're the UK? :P
<azonenberg>
Simply b/c order volume is so high the log is useless
<azonenberg>
They dont do that in the UK, do they?
<azonenberg>
for acetone?
<azonenberg>
if its some unusual hazmat i can understand logging
<sorear>
why do you think they're pushing "big data" so hard these days? need innovation to process the surveillance haul
<azonenberg>
if somebody buys pure KCN i *want* records of that :p
<rqou>
there were rumors that the uk wanted to ban acetone after some of the recent terrorist attacks, and sulfuric acid after some of the recent acid attacks
<azonenberg>
lol
<azonenberg>
well this is the same UK that, after largely banning private firearm ownership
<rqou>
of course this is unlikely to work because sulfuric acid is one of the most widely-used industrial chemicals in the world
<azonenberg>
is trying to outlaw long kitchen knives
<azonenberg>
this was years ago and i dont think it passed
<sorear>
making TATP is a self-punishing activity and criminalizing it is just cruel
<azonenberg>
but there was an attempt
<azonenberg>
sorear: hey, you get to meet your 47 virgins all right
<azonenberg>
It just probably will be long before you get to the target :p
<azonenberg>
The more terrorists we have making TATP the better
<azonenberg>
because that means they're not making something more stable and effective
<rqou>
azonenberg: also, 6+ months later i finally moved the xc2bit tests, happy? :P
<azonenberg>
lol
<azonenberg>
better late than never
<azonenberg>
rqou: how far along is your whole par at this point?
<rqou>
azonenberg: needs a lot more tests and documentation on the PAR side but it should "work"
<rqou>
needs fixing on the logic optimization side
<azonenberg>
Nice
<azonenberg>
Do you have docs etc up somewhere in the repo?
<azonenberg>
i need to try it some time
<rqou>
but i haven't tested anything beyond a trivial example and it's almost guaranteed to hit a bug
<rqou>
the last yak i was working on was to make it easier to load internal data structures
<rqou>
so that it's easier to test the corner cases i know about
<rqou>
also, no docs :P
<rqou>
who needs those :P :P
massi has joined ##openfpga
pie__ has quit [Ping timeout: 240 seconds]
m_t has joined ##openfpga
forrestv has joined ##openfpga
eduardo__ has joined ##openfpga
eduardo_ has quit [Ping timeout: 256 seconds]
m_t has quit [Remote host closed the connection]
m_t has joined ##openfpga
<q3k>
does anyone here have experience with reverse engineering IP core parameter generation in Lattice Diamond?
<q3k>
I'm trying to find out how PLL parameters are generated based on desired frequencies and phases, and it's difficult to even find the part of the suite responsible for that.
<q3k>
(I'm working on the Lattice ECP5 series)
<q3k>
okay, looks like the low level tool is called 'suba', which then loads encrypted/compressed data files from diamond/3.9_x64/ispfpga/${arch}/data/{scuba.dat,...}
<q3k>
and uses that to generate verilog wrapper, EDIFs, or whatever else
<azonenberg>
q3k: I do not have experience with that, honestly i'd rebuild it from scratch
<azonenberg>
Finding the set of legal values for a given input/output is easy
<azonenberg>
Then you just need a cost function to decide which of the possible options is best
<azonenberg>
this just requires some googling and/or testing of parameters to see if you're better off with e.g. higher, lower, or midrange VCO frequency
<azonenberg>
or higher/lower/midrange PFD
<azonenberg>
etc
massi has quit [Read error: Connection reset by peer]
azonenberg_work has quit [Ping timeout: 265 seconds]
digshadow has quit [Ping timeout: 240 seconds]
pie__ has joined ##openfpga
azonenberg_work has joined ##openfpga
digshadow has joined ##openfpga
theMagnumOrange has quit [Ping timeout: 248 seconds]
theMagnumOrange has joined ##openfpga
theMagnumOrange has quit [Ping timeout: 260 seconds]
theMagnumOrange has joined ##openfpga
theMagnumOrange has quit [Ping timeout: 265 seconds]
<rqou>
q
<rqou>
q
<rqou>
q
<m_t>
rqou, it's ESC :q ;)
theMagnumOrange has joined ##openfpga
<rqou>
no, it's "xorg crashed hard"
<rqou>
but only the gpu part, not the input part
<jn>
that sounds annoying
<rqou>
welcome to the Linux desktop?
<rqou>
xorg managed to crash and misprogram the gpu, leading to an unending spew of iommu page faults
<rqou>
also, gpu crashes are pretty scary
<rqou>
how did we survive years of buggy gpus and no iommu?
nrossi has quit [Quit: Connection closed for inactivity]
<Prf_Jakob>
rqou: lots of blood sweat and tears from GPU driver writers.
<rqou>
does having an iommu really help reduce blood sweat and tears?
<Prf_Jakob>
Not that I know of, what helped was moving to PCIe.
<Prf_Jakob>
Not deadlocking the PCI bus helped with debugging
<rqou>
ah lol
<Prf_Jakob>
But there is a reason for the channel topic of #dri-devel being:
<Prf_Jakob>
<ajax> nothing involved with X should ever be unable to find a bar
<rqou>
are you willing to talk about which chipsets you wrote drivers for?
<Prf_Jakob>
Me? I did a little bit work on i915, glanced at the PowerVR SGX and Larrabe. And then worked for VMware on the virtual graphics stack.
<Prf_Jakob>
Larrabee*
<rqou>
can you confirm or deny that the powervr architecture is a ridiculously-wide vliw with a bazillion delay slots?
<Prf_Jakob>
Can do neither, worked just a tiny bit on the 2D core. And we where outside contractors who got handed a shader compiler library that we had to use.
<rqou>
ah, ok
<Prf_Jakob>
But wouldn't suprise me :)
<rqou>
btw, intel poulsbo graphics are a giant pile of shit :P
<Prf_Jakob>
That's the thing I worked on.
<Prf_Jakob>
And yes
<rqou>
i don't even understand why intel picked it
<Prf_Jakob>
But for the power usage, and for it's time it's pretty hard to beat
<rqou>
was the old not-"hd graphics" intel gma that crappy?
<Prf_Jakob>
It sat in a PC with not so much concern for power usage.
<Prf_Jakob>
Where the Atom/Pulsbo wanted to go mobil.
<Prf_Jakob>
And the amount of tricks and hoops that the SGX does to save power is far more then Intel could ever do.
<rqou>
yeah, well it's one of the most ridiculously closed and proprietary so i have no idea how it works
<Prf_Jakob>
Basicgist, it's a tiled renderer. And then a bunch of black magic.
<Prf_Jakob>
Basic gist*
<rqou>
i know the tiled renderer part
<rqou>
and basically nothing else
<Prf_Jakob>
IIRC they had a tiling format that was in morton order.
<Prf_Jakob>
Wish I could help more, but I wasn't really attached to that project.
<balrog>
Prf_Jakob: PowerVR SGX... fuuuuun (not)
<balrog>
Poulsbo is PVR I think
<Prf_Jakob>
Yupp
<Prf_Jakob>
Bastard child of Intel and PowerVR.
<rqou>
without any working 64-bit drivers
<rqou>
even on windows
<sorear>
my knowledge of modern GPUs basically stops at having combed over the Gen specs
<sorear>
i'd like to know more but so much information isn't published
<sorear>
read a neat and enlightening book 15 years ago about the then-early history of graphics accelerators
<Prf_Jakob>
Look at the AMD architecture as well
<Prf_Jakob>
Since it's all so low level you can figure out most of what the shader core does.
<balrog>
and now, intel is working with AMD instead
<sorear>
i haven't seen any evidence that they're actually discontinuing gen
<Prf_Jakob>
Intel is big and parts often don't know what the other is doing.
<rqou>
yeah, i heard intel is particularly bad at that
<Prf_Jakob>
Which why I think the whole pulsbo thing turned out so bad. The part that does open source didn't know and have a say in the gfx core they used.