channel logs for 2004 - 2010 are archived at http://tunes.org/~nef/logs/old/ ·· can't be searched

#osdev2 = #osdev @ Libera from 23may2021 to present

#osdev @ OPN/FreeNode from 3apr2001 to 23may2021

all other channels are on OPN/FreeNode from 2004 to present

http://bespin.org/~qz/search/?view=1&c=osdev&y=18&m=1&d=5

Friday, 5 January 2018

00:00:00 --- log: started osdev/18.01.05
00:03:06 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
00:03:49 --- quit: user10032 (Quit: Leaving)
00:06:07 --- quit: hmmmm (Ping timeout: 264 seconds)
00:06:11 --- quit: Darmor (Quit: Leaving)
00:08:22 --- join: eremitah_ (~int@unaffiliated/eremitah) joined #osdev
00:09:32 --- join: hmmmm (~sdfgsf@pool-72-79-165-124.sctnpa.east.verizon.net) joined #osdev
00:10:00 --- quit: Mikaku (Quit: server maintenance ...)
00:10:58 --- quit: eremitah (Ping timeout: 240 seconds)
00:10:58 --- nick: eremitah_ -> eremitah
00:14:43 --- quit: Belxjander (Ping timeout: 252 seconds)
00:15:31 --- quit: Barrett (Quit: exit)
00:19:04 --- join: Mikaku (~Mikaku@pdpc/supporter/active/mikaku) joined #osdev
00:19:46 --- quit: Uityyy (Ping timeout: 250 seconds)
00:19:51 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
00:24:50 --- quit: daniele_athome (Ping timeout: 240 seconds)
00:27:27 --- join: daniele_athome (~daniele_a@93-40-14-81.ip36.fastwebnet.it) joined #osdev
00:29:35 --- join: nzoueidi (~nzoueidi@ubuntu/member/na3il) joined #osdev
00:30:47 --- join: bm371613 (~bartek@2a02:a317:603f:9800:391:ed9f:1c09:1285) joined #osdev
00:32:24 --- quit: promach_ (Quit: WeeChat 1.4)
00:32:46 --- quit: wcstok (Read error: Connection reset by peer)
00:36:55 <doug16k> promach, you sure you don't mean gdb's layout src command?
00:44:40 --- quit: KidBeta (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
00:56:03 --- join: KidBeta (~textual@220-244-156-86.tpgi.com.au) joined #osdev
01:01:55 --- join: PerryCS (ae8ac12f@gateway/web/freenode/ip.174.138.193.47) joined #osdev
01:02:14 --- quit: PerryCS (Client Quit)
01:10:19 --- join: quc (~quc@87.116.229.180) joined #osdev
01:14:30 --- quit: osa1 (Ping timeout: 248 seconds)
01:30:42 --- join: Asu (~sdelang@AMarseille-658-1-115-201.w86-219.abo.wanadoo.fr) joined #osdev
01:33:00 <cr1901_modern> https://twitter.com/cr1901/status/948999075613085696 I guess I'll leave this question here while it's quiet/so I can close the tab :P.
01:33:00 <cr1901_modern> Clearly I have no idea what I'm talking about, but I'm guessing someone here understands why Windoze (warning!) does things this way?
01:33:00 <bslsk05> ​twitter: <cr1901> @aionescu Why _wouldn't_ it make sense to mitigate against Variant 2 cross-process w/ different security attributes? And why are security properties/attributes encoded in a token (is that token going to be used if two processes aren't talking through "proper" IPC APIs?)?
01:33:32 --- quit: KidBeta (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
01:38:12 <latentprion> http://edition.cnn.com/2018/01/04/asia/north-korea-south-korea-talks-intl/index.html
01:38:12 <bslsk05> ​edition.cnn.com: North Korea accepts offer from South to meet for peace talks - CNN
01:38:21 <latentprion> Rocket man starting to play nice
01:39:10 --- quit: adam4813 (Quit: Connection closed for inactivity)
01:40:23 --- quit: hmmmm (Remote host closed the connection)
01:42:56 --- join: KidBeta (~textual@220-244-156-86.tpgi.com.au) joined #osdev
01:47:33 --- join: xerpi (~xerpi@216.red-88-23-234.staticip.rima-tde.net) joined #osdev
01:47:33 --- join: vdamewood (~vdamewood@unaffiliated/vdamewood) joined #osdev
01:48:26 --- quit: xerpi (Remote host closed the connection)
01:48:47 --- join: xerpi (~xerpi@216.red-88-23-234.staticip.rima-tde.net) joined #osdev
01:53:28 --- quit: godel (Ping timeout: 248 seconds)
01:55:34 <bcos> https://newsroom.intel.com/wp-content/uploads/sites/11/2018/01/Intel-Analysis-of-Speculative-Execution-Side-Channels.pdf
01:55:50 --- join: zesterer (~zesterer@host86-169-20-84.range86-169.btcentralplus.com) joined #osdev
01:56:24 <bcos> ^ seems like a nice doc for getting your head around the recent problems (wish I found it earlier..)
01:56:50 --- quit: because (Ping timeout: 268 seconds)
01:57:59 --- join: alyptik (ayy@cpe-76-173-133-37.hawaii.res.rr.com) joined #osdev
01:58:45 --- join: glauxosdever (~alex@ppp-94-66-36-254.home.otenet.gr) joined #osdev
01:59:08 --- join: eremitah_ (~int@unaffiliated/eremitah) joined #osdev
02:01:58 --- quit: eremitah (Ping timeout: 248 seconds)
02:01:58 --- nick: eremitah_ -> eremitah
02:04:13 --- join: gianluca (~user@lugbari/people/gianluca) joined #osdev
02:05:10 --- quit: caen23 (Ping timeout: 248 seconds)
02:06:23 --- quit: KidBeta (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
02:07:48 --- join: KidBeta (~textual@220-244-156-86.tpgi.com.au) joined #osdev
02:14:43 --- quit: gianluca (Ping timeout: 268 seconds)
02:16:35 --- quit: Asu (Quit: Konversation terminated!)
02:17:02 --- join: Asu (~sdelang@AMarseille-658-1-115-201.w86-219.abo.wanadoo.fr) joined #osdev
02:18:33 --- join: gianluca (~user@lugbari/people/gianluca) joined #osdev
02:18:33 --- quit: Belxjander (Ping timeout: 260 seconds)
02:18:58 --- quit: Asu (Client Quit)
02:19:23 --- join: caen23 (~caen23@79.118.94.187) joined #osdev
02:19:26 --- join: Asu (~sdelang@AMarseille-658-1-115-201.w86-219.abo.wanadoo.fr) joined #osdev
02:20:57 --- quit: Asu (Client Quit)
02:21:06 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
02:21:23 --- join: Asu (~sdelang@AMarseille-658-1-115-201.w86-219.abo.wanadoo.fr) joined #osdev
02:22:50 --- quit: gianluca (Ping timeout: 240 seconds)
02:24:43 --- quit: KidBeta (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
02:31:50 --- quit: xerpi (Quit: Leaving)
02:32:23 <zesterer> This whole Intel / speculative execution shitstorm has convinced me that RISC and simplified hardware is definitely the way to go
02:35:14 <bcos> An extremely slow RISC without speculative execution, or a RISC that's actually fast enough to be useful that suffers from the same speculative execution but doesn't have the complexity to work around them?
02:36:39 <klange> computers were a mistake
02:37:26 <bcos> The entire problem is that inside every Intel CPU complex instructions are converted to "RISC micro-code' and executed by a flawed RISC CPU. Earlier CISC CPUs like 80486 that didn't have a flawed RISC CPU inside are immune to the problem. The problem is RISC CPUs!
02:37:43 <bcos> :-)
02:38:08 --- quit: robert_ (Read error: Connection reset by peer)
02:38:26 --- join: robert_ (~hellspawn@24.96.111.33) joined #osdev
02:38:26 --- quit: robert_ (Changing host)
02:38:26 --- join: robert_ (~hellspawn@objectx/robert) joined #osdev
02:39:39 <glauxosdever> What about simpler cores but CPUs with more of them? As long as most data is cached, there won't be much slowdown from multiple cores waiting to access memory.
02:40:26 <bcos> "shared nothing message passing" architectures... :-)
02:40:27 <klange> how about a thousand-core simple, slow, sequential risc?
02:41:00 <klange> all people do is web browsing these days, right? dom rendering is super parallelizable now
02:41:01 <glauxosdever> klange: That's pushing it entirely to the other end of the spectrum.
02:41:24 <bcos> klange: Banning javascript has always been a good idea..
02:41:25 <glauxosdever> Memory reads would be too slow to be useful, I think.
02:41:30 <izabera> glauxosdever: simple cores but more of them... you just described gpus
02:41:48 <klange> basically
02:41:53 <glauxosdever> I did, ok. So?
02:42:16 <izabera> that's illegal
02:42:18 <izabera> how could you
02:42:38 <glauxosdever> I don't get your joke but anyway..
02:43:54 --- join: m3nt4L (~asvos@2a02:587:a019:8300:3285:a9ff:fe8f:665d) joined #osdev
02:46:39 <geist> note that risc machines can still be OOO and speculative. high end ARM cores, for example, have the same problem
02:46:49 <geist> what you really mean is you want a dumb, 486 class machine
02:47:13 <geist> and simple, earlier, or intentionally dump (like say cortex-a5 or a53) cores are basically dumb in order risc machines
02:48:13 <izabera> dan bernstein has long been calling for a secure machine that basically behaves like a 386 with linux on it
02:48:29 <vc-01> .. or let compilers do the work ? (referring to so far infamous VLIW architectures)
02:51:44 --- join: osa1 (~omer@212.252.142.138) joined #osdev
02:51:44 --- quit: osa1 (Changing host)
02:51:44 --- join: osa1 (~omer@haskell/developer/osa1) joined #osdev
02:56:47 --- join: m_t (~m_t@p5DDA3A0D.dip0.t-ipconnect.de) joined #osdev
02:58:55 --- part: mquin left #osdev
02:58:56 --- quit: Benjojo (Ping timeout: 246 seconds)
03:00:50 --- join: Benjojo (~sid4563@srv-83.nsa.me.uk) joined #osdev
03:05:38 --- quit: osa1 (Remote host closed the connection)
03:05:55 --- join: osa1 (~omer@haskell/developer/osa1) joined #osdev
03:13:59 --- quit: Benjojo (Ping timeout: 246 seconds)
03:16:29 --- join: Benjojo (~sid4563@srv-83.nsa.me.uk) joined #osdev
03:25:38 --- quit: silas (Ping timeout: 260 seconds)
03:25:54 --- join: silas (~silas@177.104.48.3) joined #osdev
03:28:52 --- quit: oaken-source (Ping timeout: 252 seconds)
03:35:31 --- join: oaken-source (~oaken-sou@p3E9D3048.dip0.t-ipconnect.de) joined #osdev
03:37:58 --- quit: Asu (Ping timeout: 240 seconds)
03:37:58 --- quit: Celelibi (Ping timeout: 240 seconds)
03:49:02 --- quit: ahrs (Remote host closed the connection)
03:50:12 --- join: ahrs (quassel@gateway/vpn/privateinternetaccess/ahrs) joined #osdev
03:52:02 <Sjors> http://www.righto.com/2018/01/xerox-alto-zero-day-cracking-disk.html
03:52:03 <bslsk05> ​www.righto.com: Xerox Alto zero-day: cracking disk password protection on a 45 year old system
03:52:48 <Sjors> 2018 was supposed to be good, but it starts with all these security dramas
03:52:53 --- quit: bm371613 (Quit: Konversation terminated!)
03:52:58 <Sjors> now I can't even trust my xerox alto from the 70s anymore
03:53:55 --- quit: Yun (Quit: Yun)
04:00:53 --- join: GautamS (~GautamS@59.182.252.110) joined #osdev
04:01:44 --- join: Celelibi (celelibi@par69-9-88-166-81-29.fbx.proxad.net) joined #osdev
04:05:01 --- quit: raphaelsc (Remote host closed the connection)
04:06:14 --- quit: immibis (Ping timeout: 248 seconds)
04:07:18 --- join: Asu (~sdelang@92.184.97.119) joined #osdev
04:10:39 <promach> doug16k: layout src from gdb does not work nicely in minicom terminal. it just clutters the screen
04:18:01 <rain1> https://en.wikipedia.org/wiki/Kernel_page-table_isolation
04:18:02 <bslsk05> ​en.wikipedia.org: Kernel page-table isolation - Wikipedia
04:18:06 <rain1> KPTI fixes these leaks by separating user-space and kernel-space page tables entirely
04:18:08 <rain1> how does this work
04:19:05 --- join: dhoelzer (~dhoelzer@ool-4a5940e5.dyn.optonline.net) joined #osdev
04:19:52 --- quit: m_t (Remote host closed the connection)
04:20:42 --- join: Asu` (~sdelang@AMarseille-658-1-165-254.w86-198.abo.wanadoo.fr) joined #osdev
04:20:50 --- quit: Asu (Ping timeout: 256 seconds)
04:27:55 --- join: m_t (~m_t@p5DDA3A0D.dip0.t-ipconnect.de) joined #osdev
04:29:35 <szpght> when kernel switches to user mode it unmaps all it's memory (except interrupt and syscall handlers)
04:30:03 <szpght> and when interrupt or syscall occurs it maps it again
04:37:55 --- quit: GautamS (Ping timeout: 264 seconds)
04:40:18 --- quit: caen23 (Ping timeout: 260 seconds)
04:44:11 <izabera> https://i.redd.it/dwisdy3cxm8y.png
04:44:37 <lava> szpght: that's not correct
04:44:52 <lava> working principle is described here: https://gruss.cc/files/kaiser.pdf
04:45:06 <lava> look at the neat trick in fig 3
04:45:26 <lava> switching between CR3s with minimal register clobbering
04:45:44 --- quit: zaquest (Ping timeout: 248 seconds)
04:46:23 --- part: orion left #osdev
04:48:34 <szpght> clever idea
04:48:37 <szpght> thanks for correction
04:49:59 <lava> i think it's generally a good idea to do that
04:50:06 <lava> with proper CPU support performance impact will be minimal
04:50:18 <lava> and this defeats many bugs we know and bugs we don't know yet
04:50:56 <bcos> Without PCID it trashes TLBs and ruins performance
04:51:34 --- join: uvgroovy (~uvgroovy@c-65-96-163-219.hsd1.ma.comcast.net) joined #osdev
04:54:18 --- quit: Belxjander (Ping timeout: 260 seconds)
04:56:48 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
04:58:42 --- nick: immersive -> retpoline
05:00:27 --- join: dennis95 (~dennis@p50915F3E.dip0.t-ipconnect.de) joined #osdev
05:00:34 --- join: zaquest (~notzaques@5.128.210.30) joined #osdev
05:03:03 --- quit: Asu` (Ping timeout: 260 seconds)
05:07:09 --- quit: dhoelzer (Ping timeout: 252 seconds)
05:12:12 --- join: bm371613 (~bartek@2a02:a317:603f:9800:391:ed9f:1c09:1285) joined #osdev
05:13:50 --- quit: Belxjander (Ping timeout: 240 seconds)
05:13:55 --- join: gareppa (~gareppa@unaffiliated/gareppa) joined #osdev
05:14:35 --- quit: gareppa (Remote host closed the connection)
05:15:19 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
05:20:02 --- quit: zesterer (Remote host closed the connection)
05:24:22 --- quit: awang_ (Ping timeout: 252 seconds)
05:27:04 --- quit: ahrs (Remote host closed the connection)
05:28:13 --- join: ahrs (quassel@gateway/vpn/privateinternetaccess/ahrs) joined #osdev
05:34:16 --- join: tlaxkit (~hexchat@92.189.36.210) joined #osdev
05:34:51 --- quit: m3nt4L (Remote host closed the connection)
05:39:07 --- join: Asu (~sdelang@AMarseille-658-1-101-240.w86-219.abo.wanadoo.fr) joined #osdev
05:41:26 --- join: garit2 (~garit@unaffiliated/garit) joined #osdev
05:41:29 --- quit: garit2 (Excess Flood)
05:42:46 --- quit: garit (Ping timeout: 248 seconds)
05:45:10 --- join: awang (~awang@rrcs-24-106-163-46.central.biz.rr.com) joined #osdev
05:45:34 --- quit: Humble (Ping timeout: 240 seconds)
05:47:13 --- quit: return0e (Read error: Connection reset by peer)
05:47:27 --- quit: bauen1 (Read error: Connection reset by peer)
05:47:31 --- join: qeos (~qeos@ppp158-255-171-31.pppoe.spdop.ru) joined #osdev
05:47:39 --- join: bauen1 (~bauen1@ip5f5bfcbd.dynamic.kabel-deutschland.de) joined #osdev
05:48:02 --- join: return0e (~return0e@87-102-105-145.static.kcom.net.uk) joined #osdev
05:48:38 --- quit: mniip (Ping timeout: 248 seconds)
05:48:38 --- quit: thinkpol (Ping timeout: 248 seconds)
05:48:45 --- quit: mhi^ (Read error: Connection reset by peer)
05:49:29 --- quit: lucebac (Remote host closed the connection)
05:50:14 --- quit: qeos|2 (Ping timeout: 248 seconds)
05:50:26 --- join: mhi^ (~mhi^@unaffiliated/mhi/x-9993184) joined #osdev
05:51:13 --- join: mniip (mniip@unaffiliated/mniip) joined #osdev
05:51:18 --- quit: abraxis (Ping timeout: 248 seconds)
05:51:42 --- join: abraxis (~abraxis@197.184.61.15) joined #osdev
05:52:13 --- join: lucebac (~lucebac@lucebac.net) joined #osdev
05:52:13 --- quit: lucebac (Changing host)
05:52:13 --- join: lucebac (~lucebac@unaffiliated/lucebac) joined #osdev
05:52:51 --- join: paranoidandroid_ (4ea247be@gateway/web/freenode/ip.78.162.71.190) joined #osdev
05:53:27 <paranoidandroid_> are rams byte adressable?
05:53:45 --- quit: z0ttel_ (Remote host closed the connection)
05:54:08 --- nick: multi_io_ -> multi_io
05:56:36 <lava> bcos: depends
05:56:43 <lava> bcos: we had quite good results on skylake
05:56:52 <lava> it seems they do some tagging even without PCIDs
05:56:55 <lava> not sure what though
05:57:07 <lava> on older hardware it's horrible though
05:57:10 <lava> even with PCIDs
05:57:17 <lava> paranoidandroid_: with meltdown?
05:57:33 <lava> we dump RAM at 0.5 megabyte per second on one machine
05:57:38 <lava> on others it's slower
05:58:16 <Brnocrist> skylake has PCIDs
06:00:52 --- join: ekr_ (~ekr@188.24.103.236) joined #osdev
06:01:00 <paranoidandroid_> lava: i dont know what that is, i just wondered do processors access memory one byte at a time
06:01:01 --- quit: ekr (Read error: Connection reset by peer)
06:01:32 <lava> Brnocrist: but even without setting up PCIDs we did not see a significant performance impact
06:01:42 <lava> on skylake
06:02:09 <Brnocrist> oh it is good
06:02:22 --- join: caen23 (~caen23@79.118.94.187) joined #osdev
06:06:52 --- join: garit (~garit@94.197.121.46.threembb.co.uk) joined #osdev
06:06:52 --- quit: garit (Changing host)
06:06:52 --- join: garit (~garit@unaffiliated/garit) joined #osdev
06:06:54 --- quit: garit (Excess Flood)
06:11:48 --- join: z0ttel (~z0ttel@reuenthal.z0ttel.com) joined #osdev
06:13:15 --- join: brkt (~quassel@177.69.179.225) joined #osdev
06:19:22 --- quit: Belxjander (Ping timeout: 252 seconds)
06:22:06 <nsh> is physical memory mapped from a fixed (at boot?) offset in the linux kernel's memory mapping?
06:23:19 <nsh> oh it's in dmesg mebbes
06:23:31 --- quit: sham1 (Remote host closed the connection)
06:26:09 --- join: fujisan (uid4207@gateway/web/irccloud.com/x-gpnfjsnwiipidvfy) joined #osdev
06:27:24 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
06:30:00 <nsh> hmm, need PAGE_OFFSET
06:32:59 --- join: navidr (uid112413@gateway/web/irccloud.com/x-xxskgxrutsjmxfvy) joined #osdev
06:36:29 --- join: mmu_man (~revol@vaf26-2-82-244-111-82.fbx.proxad.net) joined #osdev
06:42:02 --- join: John___ (~John__@79.97.140.214) joined #osdev
06:42:57 --- quit: uvgroovy (Quit: uvgroovy)
06:44:51 --- quit: brynet (Quit: leaving)
06:49:18 --- join: NotSecwitter (~NonSecwit@unaffiliated/nonsecwitter) joined #osdev
06:49:20 --- part: rubenwardy left #osdev
06:52:58 --- quit: mmu_man (Ping timeout: 240 seconds)
07:01:04 --- join: AndrewBC (~AndrewBC@ip70-187-57-150.pn.at.cox.net) joined #osdev
07:05:31 --- join: garit (~garit@94.197.121.46.threembb.co.uk) joined #osdev
07:05:32 --- quit: garit (Changing host)
07:05:32 --- join: garit (~garit@unaffiliated/garit) joined #osdev
07:08:38 --- quit: rain1 (Quit: WeeChat 2.0.1)
07:12:37 <lava> today it's randomized
07:12:47 <lava> but using meltdown for derandomization takes a second
07:15:03 --- join: uvgroovy (~uvgroovy@199.188.233.130) joined #osdev
07:18:30 --- join: rain1 (~user@unaffiliated/rain1) joined #osdev
07:24:39 <bcos> lava: Just need to re-randomise kernel space every 500 ms! :-)
07:24:46 --- join: uvgroovy_ (~uvgroovy@199.188.233.130) joined #osdev
07:25:14 --- join: GautamS (~GautamS@59.182.254.16) joined #osdev
07:26:01 <Brnocrist> remap kernel pags every 30ms!
07:26:33 --- quit: uvgroovy (Ping timeout: 260 seconds)
07:26:33 --- nick: uvgroovy_ -> uvgroovy
07:26:53 <rain1> hi
07:26:57 <lava> ^^
07:27:13 <lava> i'm super happy that all of you now must implement my stupid kaiser idea :D
07:27:23 <Brnocrist> lol
07:27:51 <rain1> you came up with kaiser?
07:28:24 * bcos is thinking of doing a "4 G/4 G split" thing instead (or the equivelent for 64-bit - too lazy to calculate)
07:29:50 --- quit: azonenberg_work (Ping timeout: 265 seconds)
07:32:28 <rain1> https://github.com/raphaelsc/Am-I-affected-by-Meltdown/blob/master/meltdown_checker.cc
07:32:29 <bslsk05> ​github.com: Am-I-affected-by-Meltdown/meltdown_checker.cc at master · raphaelsc/Am-I-affected-by-Meltdown · GitHub
07:33:49 --- join: hmmmm (~sdfgsf@pool-72-79-165-124.sctnpa.east.verizon.net) joined #osdev
07:34:30 <lucebac> i didnt see the `connection' between lava and meltdown/spectre/[...] until like 2h ago :D
07:34:51 <lava> :D
07:44:42 --- join: brynet (~brynet@brynet6.biz.tm) joined #osdev
07:45:14 <rain1> can someone please explain to me how kernel page table isolation works
07:47:35 <mlq> rain1: you can checkout the paper 'kaslr is dead: long live kaslr' which explains how kaiser works
07:47:45 <rain1> thank you
07:48:03 <mlq> youre welcome
07:49:38 <rain1> Gruss et al. [6] and Jang et al. [10] proposed to unmap the kernel address
07:49:41 <rain1> space in the user space and vice versa.
07:49:43 <rain1> isn't that normal
07:50:00 <rain1> what does linux normally do?
07:51:01 <smeso> "nothing"
07:51:10 <lucebac> mapped but unaccessible i guess
07:51:20 <rain1> mapped but unaccessible.. .hmmm
07:52:00 <rain1> is it like users and groups
07:52:06 <rain1> the kernel pages are only readable by kernel user
07:52:08 <smeso> it just let the kernel space memory mapped, at least on x86
07:52:20 <rain1> and user space pages are readably by only that process
07:52:30 <rain1> but the files/pages are all there in memory
07:53:50 <smeso> it's more like every process has its own "mount namespace" but with the same root (kernel memory) bind mounted in every process namespace
07:54:27 <rain1> gosh its hard to understand
07:55:14 --- join: freakazoid0223 (~IceChat9@pool-108-52-4-148.phlapa.fios.verizon.net) joined #osdev
07:56:03 <smeso> maybe this can help https://www.kernel.org/doc/Documentation/x86/x86_64/mm.txt
07:59:05 --- quit: vdamewood (Quit: Textual IRC Client: www.textualapp.com)
08:00:54 --- join: dhoelzer (~dhoelzer@ool-4a5940e5.dyn.optonline.net) joined #osdev
08:01:42 --- join: multi_io_ (~olaf@x4db3a60b.dyn.telefonica.de) joined #osdev
08:02:40 --- join: davxy (~davxy@unaffiliated/davxy) joined #osdev
08:02:44 --- join: adminseodwn (~archer@unaffiliated/remcwo9o) joined #osdev
08:02:57 <Mutabah> rain1: Before kpti, the kernel (and all of its heap/other memory) was mapped into every address space (in the top half/quater) - just in a way which meant that user code couldn't access it
08:03:10 --- join: gdh (~gdh@2605:a601:639:2c00:ec5f:4c6e:58cc:262) joined #osdev
08:03:26 <rain1> how is it that the user code can't access the kernel pages?
08:03:46 <Mutabah> They're flagged as being "kernel only" (well, not flagged as "user accessible" on x86)
08:04:17 <Mutabah> (There's a bit in PTEs which indicates that the entry is accessible by code running at rings other than 0)
08:05:05 <Mutabah> KPTI however just maps the absolute minimum into every address space, then does a (relatively expensive) address space switch upon entry to kernel mode
08:05:31 --- quit: multi_io (Ping timeout: 264 seconds)
08:05:50 <rmf1723> Mutabah: is it rings other than 0, or rings other than 3?
08:06:15 <rmf1723> Sorry, "rings other than 0", or "ring 3"?
08:06:25 <rmf1723> I thought rings 0-2 all counted as supervisor.
08:07:01 --- join: vmlinuz (~vmlinuz@2804:431:f724:dfcf:c143:4fa2:479b:571a) joined #osdev
08:07:01 --- quit: vmlinuz (Changing host)
08:07:01 --- join: vmlinuz (~vmlinuz@unaffiliated/vmlinuz) joined #osdev
08:07:01 <rain1> I see!
08:07:02 <Mutabah> rmf1723: Actually... I'm not sure :) I tend to assume 1/2 don't exist
08:07:18 --- join: darkavengr (~darkaveng@82-69-52-137.dsl.in-addr.zen.co.uk) joined #osdev
08:09:14 --- quit: oaken-source (Ping timeout: 240 seconds)
08:09:14 --- quit: davxy (Quit: davxy)
08:09:18 <darkavengr> hello
08:09:36 <rmf1723> IIRC this is one of the many reasons they're pointless: they get access to all memory.
08:10:27 --- join: davxy (~davxy@unaffiliated/davxy) joined #osdev
08:12:51 --- join: uvgroovy_ (~uvgroovy@199.188.233.130) joined #osdev
08:14:19 --- quit: glenda (Ping timeout: 252 seconds)
08:14:20 --- join: grzesiek (~grzesiek@PC-77-46-101-67.euro-net.pl) joined #osdev
08:14:58 --- quit: uvgroovy (Ping timeout: 260 seconds)
08:14:58 --- nick: uvgroovy_ -> uvgroovy
08:15:03 --- join: glenda (~glenda@mischief-1-pt.tunnel.tserv29.fmt1.ipv6.he.net) joined #osdev
08:19:23 --- join: pictron (~tom@pool-173-79-33-247.washdc.fios.verizon.net) joined #osdev
08:19:33 --- join: sham1 (~sham1@212-146-44-37.aboa.net) joined #osdev
08:20:48 --- quit: uvgroovy (Ping timeout: 260 seconds)
08:21:52 --- quit: bm371613 (Quit: Konversation terminated!)
08:26:33 --- join: sortaddsort (~archer@unaffiliated/remcwo9o) joined #osdev
08:28:30 <smeso> I think you can "hide" parts of kernel memory from them, they shouldn't be able to touch CR3
08:29:10 --- quit: adminseodwn (Ping timeout: 248 seconds)
08:29:46 <smeso> anyhow that would be painful
08:29:58 <sortaddsort> smeso, what would be painful?
08:30:02 <MrOlsen> cr3 can't be touched from any ring but 0
08:30:04 <MrOlsen> right?
08:30:13 <rain1> you're a big guy
08:30:54 <smeso> sortaddsort: using ring 1/2 to run "less privileged" code, if you have to mangle with memory mapping every time
08:30:58 --- join: mmu_man (~revol@vaf26-2-82-244-111-82.fbx.proxad.net) joined #osdev
08:33:01 --- join: sbahra (~sbahra@pool-100-2-212-39.nycmny.fios.verizon.net) joined #osdev
08:41:07 --- join: vaibhav (~vnagare@125.16.97.113) joined #osdev
08:42:28 --- join: azonenberg_work (~azonenber@172.58.43.59) joined #osdev
08:45:22 <sortaddsort> smeso, I see
08:47:27 <lava> rain1: just click through the figures in the kaiser paper: https://gruss.cc/files/kaiser.pdf
08:49:21 --- quit: AntiSpamMeta (Excess Flood)
08:49:52 --- join: AntiSpamMeta (asm@AntiSpamMeta/.) joined #osdev
08:53:12 --- nick: vaibhav -> vaibhav|afk
08:55:05 <MrOlsen> any of you guys using ELF and dynamic linking?
08:55:05 --- quit: xenos1984 (Quit: Leaving.)
08:57:23 --- join: GautamS_ (~GautamS@59.182.254.16) joined #osdev
08:59:17 --- quit: lucebac (Quit: No Ping reply in 180 seconds.)
09:01:01 --- join: Uityyy (~yaaic@2607:fb90:20de:86ce:3b:9ad1:bf82:74e1) joined #osdev
09:01:24 --- join: lucebac (~lucebac@lucebac.net) joined #osdev
09:01:24 --- quit: lucebac (Changing host)
09:01:24 --- join: lucebac (~lucebac@unaffiliated/lucebac) joined #osdev
09:01:38 --- quit: GautamS (Ping timeout: 260 seconds)
09:15:41 --- quit: azonenberg_work (Ping timeout: 265 seconds)
09:16:20 --- join: wcstok (~Me@71.197.192.147) joined #osdev
09:17:26 --- join: xenos1984 (~xenos1984@22-164-191-90.dyn.estpak.ee) joined #osdev
09:18:38 --- quit: GautamS_ (Quit: Leaving)
09:18:45 --- quit: John___ (Read error: Connection reset by peer)
09:23:43 --- join: srjek (~srjek@2601:249:601:9e9d:94fe:73c1:f28f:fe37) joined #osdev
09:31:27 --- quit: sortaddsort (Remote host closed the connection)
09:31:48 --- quit: transistor (Ping timeout: 276 seconds)
09:32:43 --- quit: darkavengr (Quit: Ex-Chat)
09:33:07 --- quit: Love4Boobies (Ping timeout: 264 seconds)
09:41:30 --- quit: Defcronyke (Read error: Connection reset by peer)
09:42:58 --- quit: dustinm` (Quit: Leaving)
09:46:18 --- join: Defcronyke (~Defcronyk@cryptospread.com) joined #osdev
09:46:39 --- join: s0n1cB00m (~s0n1cB00m@cpc83309-brig21-2-0-cust565.3-3.cable.virginm.net) joined #osdev
09:47:11 --- join: Me_ (~Me@c-71-197-192-147.hsd1.wa.comcast.net) joined #osdev
09:47:30 <rmf1723> smeso: without touching cr3 you can still change the page tables if there is e.g. a fractal mapping
09:47:34 --- join: dustinm` (~dustinm@68.ip-149-56-14.net) joined #osdev
09:47:35 --- join: azonenberg_work (~azonenber@74.85.93.91) joined #osdev
09:48:52 <smeso> rmf1723: of course you'll need to be careful
09:49:30 --- quit: wcstok (Ping timeout: 256 seconds)
09:50:47 --- nick: Me_ -> wcstok
10:00:54 --- quit: NotSecwitter (Ping timeout: 248 seconds)
10:06:08 --- quit: marcan (Quit: Now where's my screwdriver...)
10:06:08 --- join: oaken-source (~oaken-sou@p3E9D3048.dip0.t-ipconnect.de) joined #osdev
10:06:08 --- quit: Belxjander (Ping timeout: 268 seconds)
10:08:58 --- join: transistor (~trans@S01060018f8f95df7.vc.shawcable.net) joined #osdev
10:09:35 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
10:21:20 --- join: marcan (marcan@marcansoft.com) joined #osdev
10:21:30 --- quit: navidr (Quit: Connection closed for inactivity)
10:22:33 --- join: dbittman (~dbittman@2601:647:ca00:1651:b26e:bfff:fe31:5ba2) joined #osdev
10:24:56 --- quit: daniele_athome (Ping timeout: 248 seconds)
10:25:28 --- quit: brkt (Ping timeout: 240 seconds)
10:26:27 --- join: daniele_athome (~daniele_a@93-40-14-81.ip36.fastwebnet.it) joined #osdev
10:28:41 --- join: freakazoid0223_ (~IceChat9@pool-108-52-4-148.phlapa.fios.verizon.net) joined #osdev
10:29:13 --- join: NotSecwitter (~NonSecwit@unaffiliated/nonsecwitter) joined #osdev
10:29:18 --- join: Hagaric (~textual@c208F3E56.dhcp.as2116.net) joined #osdev
10:29:43 --- quit: freakazoid0223 (Ping timeout: 260 seconds)
10:31:34 --- quit: marcan (Ping timeout: 265 seconds)
10:31:45 --- join: marcan (marcan@marcansoft.com) joined #osdev
10:34:36 --- quit: dbittman (Remote host closed the connection)
10:36:16 --- quit: wcstok (Read error: Connection reset by peer)
10:38:50 --- quit: mmu_man (Ping timeout: 240 seconds)
10:45:25 --- join: dbittman (~dbittman@2601:647:ca00:1651:b26e:bfff:fe31:5ba2) joined #osdev
10:48:51 <Kazinsal> A discussion at work about the performance impact of Meltdown mitigations has turned into a pun fight
10:48:54 --- quit: Uityyy (Ping timeout: 240 seconds)
10:49:15 <Kazinsal> Someone just made a Mach pun.
10:49:16 --- join: Pessimist (~Pessimist@unaffiliated/pessimist) joined #osdev
10:49:26 <Kazinsal> I am struggling to come up with a play on words involving Minix, god dammit...
10:51:17 <sham1> IME
10:51:20 <sham1> That's your pun
10:51:21 --- join: Love4Boobies (~L4B@unaffiliated/l4b) joined #osdev
10:54:01 <geist> man i hurd that
10:55:54 <sham1> Well, let's not forget that another word for spectre is "ghost"
10:56:59 <geist> are you Mach ing me?
10:57:15 <radens> oh god that's terrible
10:58:27 <sham1> Yes. Although, I laughed more than I probably should have
11:01:10 --- quit: Belxjander (Ping timeout: 248 seconds)
11:02:57 --- quit: grzesiek (Remote host closed the connection)
11:03:43 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
11:04:20 --- quit: regreg_ (Ping timeout: 240 seconds)
11:04:52 --- quit: caen23 (Ping timeout: 256 seconds)
11:09:39 --- join: caen23 (~caen23@79.118.94.187) joined #osdev
11:12:31 --- join: regreg_ (~regreg@85.121.54.224) joined #osdev
11:13:02 --- join: mmu_man (~revol@ip-117.net-89-2-152.rev.numericable.fr) joined #osdev
11:16:29 --- join: John___ (~John__@79.97.140.214) joined #osdev
11:19:54 --- quit: srjek (Read error: Connection reset by peer)
11:22:13 --- quit: gdh (Quit: Leaving)
11:26:46 --- quit: NotSecwitter (Ping timeout: 248 seconds)
11:33:34 --- quit: John___ (Remote host closed the connection)
11:33:55 --- join: rubenwardy (~rubenward@unaffiliated/rubenwardy) joined #osdev
11:37:15 --- join: banisterfiend (~banister@ruby/staff/banisterfiend) joined #osdev
11:37:26 --- join: NotSecwitter (~NonSecwit@unaffiliated/nonsecwitter) joined #osdev
11:42:45 --- quit: vmlinuz (Ping timeout: 255 seconds)
11:48:57 <Brnocrist> https://capsule8.com/blog/detecting-meltdown-using-capsule8/
11:48:57 <bslsk05> ​capsule8.com: Detecting Meltdown using Capsule8 - Capsule8
11:49:03 <Brnocrist> this way is very very weak
11:50:23 <Brnocrist> lava, how effective is PMC to detect spectre?
11:55:29 --- quit: CompanionCube (Ping timeout: 246 seconds)
11:56:55 --- join: heat (~heat@sortix/contributor/heat) joined #osdev
11:57:03 --- join: CompanionCube (~samis@unaffiliated/drmushroom) joined #osdev
11:58:32 <Kazinsal> One of my coworkers is building a smart flow sensor for our break room to post in Spark when the carafe reaches two cups left
11:59:16 <Kazinsal> If someone takes the last cup and doesn't brew a new pot it publicly shames them
11:59:39 --- join: vmlinuz (~vmlinuz@unaffiliated/vmlinuz) joined #osdev
12:00:15 <geist> so then the questino is that is the minimum amount of coffee you can leave in the pot and not trigger it
12:00:22 <geist> like, half a cup, that's always fun
12:00:32 <geist> especially when it sits there for hours cooking the coffee
12:00:50 --- quit: vmlinuz (Remote host closed the connection)
12:01:33 <Kazinsal> He's in the lab with a soldering iron
12:01:36 <Kazinsal> Things are getting intense
12:03:44 <bcos> Just bring a thermos to use as a coffee cache - "Oh, there's always 1 cup left when I'm finished.."
12:07:26 <geist> oooh just have black water in it, pour it into the carafe before you put it back
12:09:50 --- quit: tlaxkit (Ping timeout: 240 seconds)
12:10:49 --- join: aleamb (~aleamb@unaffiliated/malpas) joined #osdev
12:14:19 --- join: jakogut (~jakogut_@162.251.69.147) joined #osdev
12:15:49 --- quit: Love4Boobies (Quit: Leaving)
12:19:16 <Kazinsal> The espresso drinkers are all going "well this isn't going to affect me"
12:25:58 --- quit: daniele_athome (Ping timeout: 248 seconds)
12:27:41 --- join: daniele_athome (~daniele_a@93-40-14-81.ip36.fastwebnet.it) joined #osdev
12:28:56 <azonenberg_work> lol
12:29:08 <azonenberg_work> espresso? you dont need a fancy coffeepot for that
12:29:15 --- quit: Pessimist (Ping timeout: 276 seconds)
12:29:16 <azonenberg_work> just take that half cup nobody drank
12:29:18 <azonenberg_work> and boil it down :P
12:30:07 --- quit: Belxjander (Ping timeout: 264 seconds)
12:31:50 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
12:33:03 --- join: Pessimist (~Pessimist@unaffiliated/pessimist) joined #osdev
12:37:52 --- quit: paranoidandroid_ (Ping timeout: 260 seconds)
12:39:30 --- join: godel (~gonzalo@96-251-231-201.fibertel.com.ar) joined #osdev
12:46:46 --- quit: caen23 (Ping timeout: 248 seconds)
12:54:38 --- join: uvgroovy (~uvgroovy@199.188.233.130) joined #osdev
12:57:14 --- join: tacco\unfoog (~tacco@dslb-188-110-204-223.188.110.pools.vodafone-ip.de) joined #osdev
12:59:24 <Kazinsal> We have an espresso machine that the espresso elitists use
13:03:44 <sham1> You should call it "expresso" to annoy them
13:12:02 --- quit: [Brain] (Ping timeout: 265 seconds)
13:12:53 --- join: [Brain] (~brain@brainwave.brainbox.cc) joined #osdev
13:21:20 --- quit: godel (Ping timeout: 268 seconds)
13:22:42 --- quit: abraxis (Remote host closed the connection)
13:30:54 --- quit: nzoueidi (Ping timeout: 240 seconds)
13:32:12 --- quit: sigjuice (Ping timeout: 256 seconds)
13:33:29 --- join: sigjuice (~sigjuice@107.170.193.86) joined #osdev
13:45:14 --- quit: AnyTimeTraveler (Read error: Connection reset by peer)
13:46:50 --- join: Shamar (~giacomote@unaffiliated/giacomotesio) joined #osdev
13:49:34 --- quit: Asu (Ping timeout: 240 seconds)
13:49:44 --- join: Asu (~sdelang@92.184.100.251) joined #osdev
13:49:59 --- join: Sasha- (0226cdf7@gateway/web/cgi-irc/kiwiirc.com/ip.2.38.205.247) joined #osdev
13:50:14 --- part: Sasha- left #osdev
13:53:41 --- join: atrapado_ (~atrapado@unaffiliated/atrapado) joined #osdev
14:02:28 --- quit: dennis95 (Quit: Leaving)
14:20:43 --- quit: NotSecwitter (Ping timeout: 260 seconds)
14:26:29 --- join: CheckDavid (uid14990@gateway/web/irccloud.com/x-qxdgkezmgydtrxni) joined #osdev
14:30:01 --- join: xerpi (~xerpi@216.red-88-23-234.staticip.rima-tde.net) joined #osdev
14:30:04 --- quit: heat (Ping timeout: 255 seconds)
14:32:18 --- quit: glauxosdever (Quit: leaving)
14:33:20 --- quit: oaken-source (Ping timeout: 240 seconds)
14:40:10 --- quit: return0e (Remote host closed the connection)
14:41:19 --- quit: lucebac (Quit: No Ping reply in 180 seconds.)
14:41:26 --- join: lucebac (~lucebac@lucebac.net) joined #osdev
14:41:26 --- quit: lucebac (Changing host)
14:41:26 --- join: lucebac (~lucebac@unaffiliated/lucebac) joined #osdev
14:51:56 --- quit: atrapado_ (Quit: Leaving)
14:58:30 --- join: sortie (~sortie@static-5-186-55-44.ip.fibianet.dk) joined #osdev
14:59:46 --- join: thinkpol (~thinkpol@130.225.0.251) joined #osdev
15:08:28 --- quit: banisterfiend (Ping timeout: 252 seconds)
15:11:10 --- join: freakazoid0223 (~IceChat9@pool-108-52-4-148.phlapa.fios.verizon.net) joined #osdev
15:11:32 --- join: heat (~heat@sortix/contributor/heat) joined #osdev
15:12:54 --- quit: freakazoid0223_ (Ping timeout: 248 seconds)
15:13:26 --- join: AnyTimeTraveler (~quassel@titan-systems.net) joined #osdev
15:13:42 --- join: immibis (~chatzilla@122-59-200-50.jetstream.xtra.co.nz) joined #osdev
15:14:15 --- quit: fujisan (Quit: Connection closed for inactivity)
15:14:51 --- quit: aleamb (Remote host closed the connection)
15:14:58 --- quit: cr1901_modern (Read error: Connection reset by peer)
15:37:24 --- quit: m_t (Quit: Leaving)
15:38:43 --- join: raphaelsc (~utroz@179.186.168.22.dynamic.adsl.gvt.net.br) joined #osdev
15:39:09 <raphaelsc> guys, is there a way to know the range of kernel address space if my program was compiled with -fPIC?
15:39:36 <raphaelsc> i may assume it's static and depends on the kernel version, right?
15:39:43 <heat> raphaelsc: Huh?
15:39:48 <heat> It's always the same?
15:40:21 <raphaelsc> heat: with position independent flag, perhaps. does it override ASLR?
15:40:36 <heat> what do you mean with "Kernel address space range"?
15:40:49 --- join: _sfiguser (~sfigguser@host194-66-dynamic.116-80-r.retail.telecomitalia.it) joined #osdev
15:40:49 <heat> ASLR has nothing to do with it
15:40:59 <heat> neither does PIC
15:41:09 <raphaelsc> heat: where the kernel was mapped in the the address space of the user process.
15:41:24 <raphaelsc> [start, start + len]
15:41:24 <heat> the kernel image or the kernel's memory?
15:41:43 <raphaelsc> both, preferably the latter.
15:41:47 --- quit: _sfiguser (Client Quit)
15:42:05 <heat> the kernel image is randomized but it's between 0xffffffff80000000 and 0xffffffffffffffff
15:42:22 <raphaelsc> thanks! what about kernel's memory?
15:42:39 <heat> the kernel's memory is between 0xffff800000000000 and 0xffffffffffffffff
15:43:04 <heat> although in documentation/x86_64 there's a neat .txt describing the individual memory ranges
15:43:21 <raphaelsc> i'll take a look. thanks!
15:43:26 <heat> and the range might change between cpus since the virtual address space bits might differ
15:44:17 <heat> raphaelsc: Oh, and without KASLR the kernel image starts at 0xffffffff80000000 + physical base address(defaults to 16MiB, or 0x1000000)
15:44:53 <heat> and you just won't find it mapped if you're running linux with PTI
15:45:15 --- join: KidBeta (~textual@220-244-156-86.tpgi.com.au) joined #osdev
15:45:28 --- quit: KidBeta (Max SendQ exceeded)
15:47:06 <raphaelsc> heat: yeah know that.
15:47:40 --- quit: Asu (Quit: Konversation terminated!)
15:48:19 --- join: navidr (uid112413@gateway/web/irccloud.com/x-vgenstcranapjfka) joined #osdev
15:49:20 --- join: freakazoid0223_ (~IceChat9@pool-108-52-4-148.phlapa.fios.verizon.net) joined #osdev
15:49:54 --- quit: freakazoid0223 (Ping timeout: 256 seconds)
15:50:19 --- quit: xerpi (Quit: Leaving)
15:56:40 --- join: ebrasca (~user@host-198.58.elzappero.net) joined #osdev
16:02:48 --- quit: mmu_man (Ping timeout: 260 seconds)
16:10:55 --- quit: awang (Ping timeout: 264 seconds)
16:11:31 --- quit: regreg_ (Ping timeout: 264 seconds)
16:17:09 --- join: mmu_man (~revol@vaf26-2-82-244-111-82.fbx.proxad.net) joined #osdev
16:17:25 --- join: zwliew (uid161395@gateway/web/irccloud.com/x-mhgdvuqlzcwyqmaz) joined #osdev
16:21:04 --- quit: uvgroovy (Ping timeout: 256 seconds)
16:21:30 --- join: adam4813 (uid52180@gateway/web/irccloud.com/x-xvuzwhqjqjpszjhx) joined #osdev
16:23:43 --- join: entropy_ (185afcbf@gateway/web/freenode/ip.24.90.252.191) joined #osdev
16:23:55 <entropy_> Hi
16:24:51 <heat> hi
16:25:30 <pixelherodev> Suilod, mellon
16:25:38 <pixelherodev> Sorry, not the place for that
16:25:53 <entropy_> How can I make update-grub recognize my kernel?
16:25:53 --- join: awang (~awang@cpe-98-31-27-190.columbus.res.rr.com) joined #osdev
16:26:30 <entropy_> can I make update-grub recognize my kernel?
16:27:01 <pixelherodev> Probably. Did you Google "update-grub multiboot"?
16:27:04 <heat> entropy_: You don't
16:27:20 <heat> Not the best idea
16:30:15 <entropy_> Because I don't think it's safe to directly add a menuentry in grub.cfg, but os-prober won't recognize the kernel
16:30:42 <heat> oh it's safe
16:30:53 <heat> might not boot though
16:30:57 --- quit: jakogut (Remote host closed the connection)
16:31:16 <heat> afaik the multiboot modules aren't included by default
16:31:27 <entropy_> the computer won't boot or my kernel won't boot?
16:31:33 <heat> your kernel won't boot
16:31:47 <entropy_> ok thanks
16:32:02 <heat> the safest idea is to just burn the iso to a usb drive and boot it
16:32:29 <entropy_> yes but my computer won't recognize a few usbs
16:32:37 <entropy_> only grub would
16:34:46 --- quit: CheckDavid (Quit: Connection closed for inactivity)
16:35:40 --- join: jakogut (~jakogut_@162.251.69.147) joined #osdev
16:36:14 --- quit: entropy_ (Quit: Page closed)
16:36:17 --- quit: inode (Quit: )
16:38:50 --- quit: vaibhav|afk (Quit: Leaving)
16:39:13 --- quit: azonenberg_work (Ping timeout: 252 seconds)
16:40:56 --- quit: neakitten (Disconnected by services)
16:46:57 --- quit: sortie (Quit: Leaving)
16:48:42 <raphaelsc> does anyone know how to export a tag in asm inline? like extern char[] label; func() asm( bla bla bla label: nop );
16:49:05 --- quit: xenos1984 (Quit: Leaving.)
16:51:01 <heat> raphaelsc: .global label ?
16:51:02 --- join: return0e (~return0e@87-102-105-145.static.kcom.net.uk) joined #osdev
16:54:38 --- join: neakitten (alice@hypatia/staff/nea) joined #osdev
16:55:19 --- quit: jakogut (Ping timeout: 264 seconds)
16:56:58 <raphaelsc> heat: thanks, but still getting this Error: symbol `__speculative_byte_load_exit' is already defined
16:57:22 <heat> raphaelsc: code?
16:57:47 <raphaelsc> https://pastebin.com/t9QyVyRq
16:57:48 <bslsk05> ​pastebin.com: extern char __speculative_byte_load_exit[]; __attribute__((always_inline)) i - Pastebin.com
16:58:09 <raphaelsc> it worked when i removed __attribute__((always_inline)) :-)
16:58:27 <heat> well, of course
16:58:31 <raphaelsc> that's because inline causes redefinition of a global variable, right?
16:58:38 <heat> inline will just create various global symbols
17:00:16 --- join: jakogut (~jakogut_@162.251.69.147) joined #osdev
17:07:48 --- join: wcstok (~Me@c-71-197-192-147.hsd1.wa.comcast.net) joined #osdev
17:08:54 --- join: dwn (~cute@unaffiliated/dwn) joined #osdev
17:12:54 --- quit: awang (Ping timeout: 248 seconds)
17:27:50 --- quit: sprocklem (Ping timeout: 248 seconds)
17:30:06 --- join: sprocklem (~sprocklem@unaffiliated/sprocklem) joined #osdev
17:30:35 --- join: selfReferentialN (ad4f345d@gateway/web/freenode/ip.173.79.52.93) joined #osdev
17:30:53 --- join: awang (~awang@cpe-98-31-27-190.columbus.res.rr.com) joined #osdev
17:36:31 --- part: selfReferentialN left #osdev
17:43:59 --- join: variable (~variable@freebsd/developer/variable) joined #osdev
17:45:21 --- quit: variable (Remote host closed the connection)
17:45:54 <doug16k> raphaelsc, export a tag? what do you mean?
17:46:17 <doug16k> you mean make a label a global symbol?
17:47:02 <doug16k> if that, then you need `.global label` before the label
17:47:08 --- join: variable (~variable@freebsd/developer/variable) joined #osdev
17:47:44 <doug16k> char[] is nonsense. you must mean `char label[];`
17:48:06 <doug16k> er, extern char label[];
17:48:29 <heat> or just extern char label if you don't plan on accessing it as an array
17:49:00 <doug16k> declaring as an array makes it decay into a pointer automatically
17:49:21 <heat> really?
17:49:38 <heat> ohno why didn't I know this
17:51:49 --- quit: variable (Quit: Found 1 in /dev/zero)
17:53:54 --- quit: Kazinsal (Read error: Connection reset by peer)
17:55:23 --- join: Kazinsal (~Kazinsal@unaffiliated/blacklightos) joined #osdev
18:03:46 --- quit: _ether_ (Ping timeout: 268 seconds)
18:04:23 --- quit: dhoelzer (Ping timeout: 268 seconds)
18:12:42 --- quit: qeos (Read error: Connection reset by peer)
18:21:56 --- join: Retr0id (~Retr0id@unaffiliated/retr0id) joined #osdev
18:22:08 --- join: mischief_ (~none@iota.offblast.org) joined #osdev
18:24:12 --- join: greaser|q (greaser@antihype.space) joined #osdev
18:26:07 --- join: RudyValencia_ (rudy@unaffiliated/rudyvalencia) joined #osdev
18:26:11 --- join: heinrich5991_ (~hein5991@unaffiliated/heinrich5991) joined #osdev
18:26:14 --- join: jp| (ayy@youlosethega.me) joined #osdev
18:26:18 --- join: Xark_ (~Xark@47.150.69.188) joined #osdev
18:26:28 --- join: Beato (~Beato@unaffiliated/beato) joined #osdev
18:26:41 --- quit: zwliew (Quit: Connection closed for inactivity)
18:26:54 --- join: abra0_ (moo@unaffiliated/abra0) joined #osdev
18:27:10 --- quit: [Brain] (*.net *.split)
18:27:11 --- quit: teej (*.net *.split)
18:27:11 --- quit: doug16k (*.net *.split)
18:27:11 --- quit: babyflakes (*.net *.split)
18:27:11 --- quit: clever (*.net *.split)
18:27:12 --- quit: mischief (*.net *.split)
18:27:14 --- quit: abra0 (*.net *.split)
18:27:14 --- quit: jp (*.net *.split)
18:27:14 --- quit: GreaseMonkey (*.net *.split)
18:27:14 --- quit: heinrich5991 (*.net *.split)
18:27:14 --- quit: Wild (*.net *.split)
18:27:14 --- quit: lightwalker (*.net *.split)
18:27:14 --- quit: Xark (*.net *.split)
18:27:14 --- quit: RudyValencia (*.net *.split)
18:27:56 --- join: [Brain] (~brain@brainwave.brainbox.cc) joined #osdev
18:28:00 --- join: Wild (~Wild@archlinux/trusteduser/wild) joined #osdev
18:28:04 --- join: clever (~clever@108.175.83.38) joined #osdev
18:28:04 --- quit: clever (Changing host)
18:28:04 --- join: clever (~clever@unaffiliated/clever) joined #osdev
18:28:20 --- nick: heinrich5991_ -> heinrich5991
18:28:34 --- nick: jp| -> jp
18:28:38 --- nick: abra0_ -> abra0
18:28:38 --- quit: kasumi-owari (Ping timeout: 260 seconds)
18:29:48 --- quit: sham1 (Ping timeout: 260 seconds)
18:31:00 --- join: kasumi-owari (~kasumi-ow@ftth-213-233-237-007.solcon.nl) joined #osdev
18:31:11 --- quit: listenmore (Remote host closed the connection)
18:31:35 --- join: listenmore (~strike@2.27.123.231) joined #osdev
18:31:56 --- join: ahrs_ (quassel@gateway/vpn/privateinternetaccess/ahrs) joined #osdev
18:32:08 --- quit: ahrs (Ping timeout: 260 seconds)
18:32:20 --- join: sham1 (~sham1@212-146-44-37.aboa.net) joined #osdev
18:34:21 --- join: doug16k (~dougx@174-138-193-47.cpe.distributel.net) joined #osdev
18:37:36 --- join: teej (uid154177@gateway/web/irccloud.com/x-xrxnuhklxzgaapka) joined #osdev
18:39:07 --- join: sammwch (~heddwch@76.8.4.254) joined #osdev
18:52:25 --- quit: Shamar (Quit: leaving)
18:59:17 --- join: unixpickle (~alex@2601:645:8103:60b6:3d4f:ca12:7229:1958) joined #osdev
19:07:14 --- quit: adam4813 (Quit: Connection closed for inactivity)
19:08:37 --- join: gamozo (~pleb@96.74.23.209) joined #osdev
19:08:40 <gamozo> doo doo doo
19:14:01 --- quit: gamozo (Read error: Connection reset by peer)
19:14:16 --- join: gamozo (~pleb@96.74.23.209) joined #osdev
19:17:01 --- join: voidah (~voidah@unaffiliated/voider) joined #osdev
19:29:31 --- quit: sammwch (Ping timeout: 264 seconds)
19:29:37 --- quit: heat (Remote host closed the connection)
19:30:21 --- join: freakazoid0223 (~IceChat9@pool-108-52-4-148.phlapa.fios.verizon.net) joined #osdev
19:31:34 --- quit: freakazoid0223_ (Ping timeout: 248 seconds)
19:33:00 --- quit: Retr0id (Quit: Leaving)
19:34:12 <doug16k> I'm thinking of getting a 50" 4K monitor
19:35:59 --- join: sammwch (~heddwch@75-98-151-98.cpe.safelink.net) joined #osdev
19:36:11 --- quit: sammwch (Client Quit)
19:39:03 --- quit: greaser|q (Changing host)
19:39:03 --- join: greaser|q (greaser@unaffiliated/greasemonkey) joined #osdev
19:39:22 --- nick: greaser|q -> GreaseMonkey
19:51:18 <gamozo> eww
19:51:21 <gamozo> way too big
19:51:26 <Mutabah> That seems excessive
19:51:27 <gamozo> shit my 28" monitor gives me neck pain
19:51:36 <gamozo> i try to stick with <=24" for everything now
19:51:37 <Mutabah> I have a 27" and it's a little too big, even at a bit of a distance
19:55:56 <Kazinsal> A 48" 4K would be the same DPI as a 24" 1080p
19:56:30 <Kazinsal> It'd be like being John Carmack with his 1080p CRT in an age when everyone else was using 1024x768
20:13:10 --- quit: zng (Ping timeout: 252 seconds)
20:15:12 --- quit: heddwch (Quit: ZNC 1.6.5 - http://znc.in)
20:15:12 --- quit: shikhin (Quit: Alas.)
20:17:06 --- quit: freakazoid0223 (Quit: I used to think I was indecisive, but now I'm not too sure.)
20:18:00 --- join: babyflakes (uid171740@gateway/web/irccloud.com/x-lqvrngykerasuprm) joined #osdev
20:19:04 <doug16k> yeah I was worried about my neck
20:20:19 <doug16k> I have a 27" 1920x1080 now
20:20:50 --- quit: mmu_man (Ping timeout: 240 seconds)
20:21:41 --- join: NotSecwitter (~NonSecwit@unaffiliated/nonsecwitter) joined #osdev
20:21:46 --- join: sammwch (~heddwch@2604:180:1:10e::3e89) joined #osdev
20:22:12 --- join: shikhin (shikhin@2604:180:1:10e::3e89) joined #osdev
20:30:04 --- quit: sammwch (Quit: ZNC 1.6.5 - http://znc.in)
20:30:12 --- join: heddwch (heddwch@192.184.95.158) joined #osdev
20:30:20 --- quit: NotSecwitter (Ping timeout: 240 seconds)
20:31:46 --- quit: Kazinsal (Read error: Connection reset by peer)
20:37:26 --- join: NotSecwitter (~NonSecwit@unaffiliated/nonsecwitter) joined #osdev
20:37:38 --- quit: NotSecwitter (Read error: Connection reset by peer)
20:45:43 --- quit: garit (Ping timeout: 264 seconds)
20:45:54 --- quit: tacco\unfoog ()
20:55:57 --- join: Kazinsal (~Kazinsal@unaffiliated/blacklightos) joined #osdev
20:56:08 --- quit: navidr (Quit: Connection closed for inactivity)
21:03:59 --- quit: s0n1cB00m ()
21:04:52 --- quit: quc (Ping timeout: 252 seconds)
21:10:16 --- join: quc (~quc@host-89-230-164-119.dynamic.mm.pl) joined #osdev
21:17:40 --- join: adminseodwn (~archer@unaffiliated/remcwo9o) joined #osdev
21:23:39 --- join: promach_ (~promach@167.220.255.139) joined #osdev
21:24:38 --- quit: Belxjander (Ping timeout: 248 seconds)
21:24:41 --- quit: promach_ (Client Quit)
21:26:58 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
21:40:06 --- quit: Belxjander (Ping timeout: 248 seconds)
21:40:59 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
21:47:12 --- join: sortaddsort (~archer@unaffiliated/remcwo9o) joined #osdev
21:47:31 --- quit: adminseodwn (Ping timeout: 264 seconds)
21:51:31 <gamozo> that being said dont get a 4k unless it's pretty big
21:51:40 <gamozo> scaling applications on it just never seems to work
21:51:44 <gamozo> i really regret buying my 4k
21:51:51 <gamozo> besides games everything on it is awful
21:52:14 <gamozo> either too small to read without scaling, or with scaling everything is blurry and distorted (except for the few apps that scale correctly)
21:57:44 --- quit: daniele_athome (Ping timeout: 248 seconds)
21:58:14 --- quit: Belxjander (Ping timeout: 248 seconds)
21:58:19 --- join: caen23 (~caen23@79.118.94.187) joined #osdev
21:58:29 --- join: daniele_athome (~daniele_a@93-40-14-81.ip36.fastwebnet.it) joined #osdev
21:58:40 --- join: xenos1984 (~xenos1984@2001:bb8:2002:200:6651:6ff:fe53:a120) joined #osdev
21:59:30 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
22:03:00 --- join: variable (~variable@freebsd/developer/variable) joined #osdev
22:11:34 --- join: fodil (~fodil@2a01:e35:39c9:20c0:b173:3ef8:9715:f9cc) joined #osdev
22:13:35 --- join: fodil__ (~fodil@2a01:e35:39c9:20c0:b173:3ef8:9715:f9cc) joined #osdev
22:14:01 --- quit: fodil (Client Quit)
22:19:41 --- join: zwliew (uid161395@gateway/web/irccloud.com/x-nxkntalxdqzotpcz) joined #osdev
22:34:30 --- quit: Belxjander (Ping timeout: 248 seconds)
22:36:39 --- nick: RudyValencia_ -> RudyValencia
22:37:39 --- quit: unixpickle (Quit: My MacBook has gone to sleep. ZZZzzz…)
22:39:35 <doug16k> gamozo, yeah, that's why I was trying to go really big. 50" might be a bit ridiculous though
22:40:19 <doug16k> funny thing is, at 50", it will still have more density than my 27" 1920x1080
22:41:21 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
22:44:31 <doug16k> geordi, << 1
22:45:27 <doug16k> looks like speculative execution killed geordi D:
22:48:15 <geist> did someone SPECTRE geordi? not nice!
22:48:42 --- join: osa1_ (~omer@212.252.142.138) joined #osdev
22:48:56 --- quit: osa1_ (Remote host closed the connection)
22:54:33 --- join: osa1_ (~omer@212.252.142.138) joined #osdev
22:54:35 --- quit: osa1_ (Remote host closed the connection)
22:56:16 --- join: oaken-source (~oaken-sou@p3E9D2692.dip0.t-ipconnect.de) joined #osdev
23:08:29 --- quit: wcstok (Read error: Connection reset by peer)
23:13:00 --- join: unixpickle (~alex@2601:645:8103:60b6:bd6f:9978:c669:8297) joined #osdev
23:16:19 --- join: garit (~garit@94.197.121.46.threembb.co.uk) joined #osdev
23:16:19 --- quit: garit (Changing host)
23:16:19 --- join: garit (~garit@unaffiliated/garit) joined #osdev
23:16:41 --- quit: sortaddsort (Ping timeout: 265 seconds)
23:19:56 <doug16k> geist, yeah, Eelis took geordi down when the news came out
23:20:49 --- join: adminseodwn (~archer@unaffiliated/remcwo9o) joined #osdev
23:21:19 <doug16k> so, to have the same density as 27" HD, you'd need a 54" UHD monitor
23:21:58 --- quit: dbittman (Ping timeout: 255 seconds)
23:22:12 <variable> I am kind of annoyed at how mad people are at intel
23:22:24 <variable> I mean, yes, it is a CPU bug, but its really a new class of attacks
23:22:29 <doug16k> why wouldn't they be?
23:22:53 <doug16k> the perf gained by not checking would be practically zero
23:23:39 <doug16k> what would the time window where you would benefit be? not very long
23:23:47 <variable> doug16k: true, and I don't fault people for being annoyed
23:24:00 <variable> but in advance of this bug, would you have even thought that this was a vector for attack?
23:24:16 --- join: sortaddsort (~archer@unaffiliated/remcwo9o) joined #osdev
23:24:50 <doug16k> this attack is like all truly good ideas: obvious once you know about it
23:25:12 <variable> doug16k: its kind of like the attacks on hyperthreading from back in the day: tis a real attack, and its something to mitigate, but the question is should good engineers have thought about it
23:25:24 <variable> exactly! Its obvious in hindsight
23:25:43 <variable> but its not like someone didn't do their job, or didn't take security into account
23:26:16 <doug16k> it's not like they accidentally allowed it to bypass permission checks though. they almost certainly consciously decided to do that
23:26:32 <raphaelsc> does somebody know a way to disable pti?
23:26:49 <variable> raphaelsc: on what kernel?
23:27:11 <variable> doug16k: true.
23:27:48 --- quit: adminseodwn (Ping timeout: 265 seconds)
23:28:28 <raphaelsc> Linux raphael-instance-1 3.10.0-693.11.6.el7.x86_64 #1 SMP Thu Jan 4 01:06:37 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux
23:28:36 <doug16k> I've timed loops and inferred that it must have been going waaay out of order, several loops ahead. modern cpus get miles ahead and deeply overlap things. that is why this attack works, the sheer distance it can speculate is insane
23:28:41 <raphaelsc> variable: i think the way to go is to disable pti on boot command line
23:29:12 <variable> raphaelsc: maybe you want to ask in a Linux channel? I know little about it
23:29:13 <variable> doug16k: yeah
23:30:04 <klange> raphaelsc: nopti or pti=off on kernel command line
23:30:11 <klange> or so i heard
23:30:25 <geist> doug16k: hey you done did nvme
23:30:44 <geist> have you ever seen a device that actually supported SGLs?
23:30:45 <variable> apparently arm had undocumented speculation barriers or so I heard?
23:30:52 <variable> geist: SGL ?
23:30:55 <raphaelsc> klange: thanks :-)
23:31:03 <geist> we wrote our nvme driver for fuchsia this week, and did some checking of various hardware floating around
23:31:08 <geist> turns out none of them support GSL
23:31:10 <geist> SGL even
23:31:38 <geist> variable: scatter gather list, i believe. NVME has two formats for referring to memory
23:31:43 <variable> ah
23:31:54 <variable> didn't know the acronym
23:31:56 --- quit: unixpickle (Quit: My MacBook has gone to sleep. ZZZzzz…)
23:32:13 <geist> well, 3. either you point directly to up to 2 pages in the main command descriptor, or a pointer over to a page of memory full of page pointers, or some sort of complex scatter gather list
23:32:29 <variable> yeah, like writev/readv
23:33:20 --- quit: Belxjander (Ping timeout: 240 seconds)
23:34:32 --- join: hppavilion[1] (~dosgmowdo@58-0-174-206.gci.net) joined #osdev
23:34:35 <doug16k> geist, I didn't notice, I used PRPs only
23:34:45 <doug16k> afaik, SGL wasn't supported in the first NVME spec
23:36:04 <doug16k> my understanding is, PRP requires all pointers after the first page to be page aligned, SGL doesn't
23:36:07 --- join: Belxjander (~Belxjande@sourcemage/Mage/Abh-Elementalist) joined #osdev
23:36:24 <doug16k> in other words, PRP is contiguous, SGL allows arbitrary scatter/gather
23:36:49 <doug16k> contiguous as in, allows any starting point, but from then on, always starts at a page boundary
23:37:13 <geist> yah
23:37:27 <geist> anyway, it's pretty neat
23:37:53 <doug16k> yeah
23:49:04 <doug16k> geist, did the dev try the nvme tracing patch in qemu? qemu accepted it btw
23:50:00 <doug16k> s/the/my/ :D
23:50:27 <geist> nah, did it on real hardware
23:50:43 <variable> real hardware sucks
23:50:44 <geist> actually qemu proved to be more annoying because of the way it triggers IRQs was wonky
23:51:15 <doug16k> I actually have an issue with my NVME implementation. it works fine in qemu, I get IRQ storm on real hardware
23:51:31 <geist> i think the big difference was we were generally using MSI on real hardware
23:52:02 <variable> doug16k: have you considered getting an umbrella ?
23:52:19 <doug16k> I try to use MSI but qemu only supports pin IRQ or MSI-X, so in qemu I use pin interrupt
23:52:47 <geist> yah. we had the same thing, and the qemu pin interrupt was acting weird, but he fixed it
23:52:52 <variable> :(
23:53:01 <geist> https://fuchsia-review.googlesource.com/c/zircon/+/107058 is the driver
23:53:01 <bslsk05> ​fuchsia-review.googlesource.com <no title>
23:53:27 <doug16k> variable, I'm canadian, I can withstand storms very well, no need for umbrella
23:55:48 <geist> also oooh new Steins;Gate anime coming this year
23:55:55 <geist> the first one was a truly great show
23:59:59 --- log: ended osdev/18.01.05