02:18 <sb0> SturmFlut, what does hdparm -t say?

03:01 <GitHub88> [artiq] jboulder opened pull request #10: fix __init__.py in pdq2 and lda (master...for-upstream) http://git.io/bZJ8

03:33 <GitHub126> [artiq] jboulder opened pull request #11: Novatech (master...novatech) http://git.io/bZqs

08:19 <GitHub179> [artiq] sbourdeauducq pushed 3 new commits to master: http://git.io/bneK

08:19 <GitHub179> artiq/master 1b122dd Sebastien Bourdeauducq: sync_struct,gui: refactor Subscriber data distribution to support access to parameter DB from GUI file

08:19 <GitHub179> artiq/master 2f4a83b Sebastien Bourdeauducq: gui: access to parameter DB from GUI files

08:19 <GitHub179> artiq/master 186aa7f Sebastien Bourdeauducq: style

10:43 <GitHub181> [artiq] sbourdeauducq pushed 1 new commit to master: http://git.io/bn6r

10:43 <GitHub181> artiq/master af79392 Sebastien Bourdeauducq: master: basic dumping of results to HDF5

11:15 <GitHub146> [artiq] sbourdeauducq pushed 4 new commits to master: http://git.io/bnQG

11:15 <GitHub146> artiq/master f308bf6 Sebastien Bourdeauducq: doc/manual/installing: add python3-h5py for Xubuntu

11:15 <GitHub146> artiq/master 0e65f71 Sebastien Bourdeauducq: artiq_run: HDF5 output support

11:15 <GitHub146> artiq/master 0ba07e6 Sebastien Bourdeauducq: artiq/tools/verbosity_args: add help

13:14 <sturmflut> sb0: I get about 74.23 MB/sec with all kernels (3.17.8 vanilla, 3.16.7 vanilla, 3.16.0-23-generic Ubuntu)

13:14 <sturmflut> sb0: I found something interesting regarding the Ubuntu 3.16.0-23-generic kernel

13:15 <sb0> ah, that's different than mine. I get ~120 in HS200 (buggy) and ~40 with debug_quirks on 3.13 (reliable)

13:15 <sturmflut> sb0: 3.16.0-23-generic is based on 3.16.7 stable, so I took 3.16.7 stable from kernel.org, built it with the Ubuntu 3.16.0-23-generic kernel config. And guess what: While the Ubuntu kernel is stable, my custom 3.16.7 stable from kernel.org freezes with Timeout waiting for hardware interrupt"

13:16 <sturmflut> sb0: A diff between the Ubuntu kernel and 3.16.7 stable shows some differences regarding the MMC code

13:17 <sturmflut> sb0: http://pastebin.ubuntu.com/10053339/

13:37 <sturmflut> sb0: There's a patch to drivers/mmc/host/sdhci-pci.c, adding a quirk for Intel devices, could that be what we're looking for?

13:39 <sb0> I don't think so, afaict the SDHCI thing is not on the PCI bus

13:39 <sb0> is it?

13:42 <sturmflut> At least it isn't clearly labeled on http://images.anandtech.com/doci/7314/Screen%20Shot%202013-09-11%20at%208.27.39%20AM.png

13:43 <sb0> does it show in your lspci? iirc there was only USB and video in mine

13:43 <sb0> and I don't remember the system loading sdhci_pci, it used sdhci_acpi instead

13:45 <sturmflut> You're right, it's sdhci_acpi

13:47 <sturmflut> But then where's the difference, why does the Ubuntu kernel work and the stable kernel not

13:48 <sb0> sdhci.c is also changed

13:48 <sb0> iirc

13:51 <sturmflut> No, at least I don't see it being changed

13:56 <sturmflut> sb0: On the WiFi issue, if I pass "debug=0x1416" to brcmfmac I don't get any debug output. I rebuilt the kernel with CONFIG_BRCM_TRACING=y and CONFIG_BRCMDBG=y, do I need to set more?

13:58 <sb0> I think those are the options you need

22:40 <mog> does anyone have any advice about building an lm32 toolchain from scratch?

22:40 <mog> it seems like only old versions of gcc are supported?

23:07 <sb0> no, latest works

23:07 <mog> can someone show me how to build a tool chain?