users

users@lists.genode.org

6 participants
2306 discussions

libc lwip recv MSG_PEEK doesn't behave as specified
by Johannes Kliemann 20 Jun '18

20 Jun '18

Hi all, I'm currently trying to use a recv call on a network socket with the MSG_PEEK flag. According to the man 2 recv: > MSG_PEEK causes the receive operation to return data from the beginning of the receive queue without removing that data from the queue. Thus, a subsequent receive call will return the same data. Yet this doesn't seem to be the case (at least with lwip). I implemented a test [1] that creates a connection from the test http client and subsequently calls recv(sock, buf, 1, MSG_PEEK) which always should yield 'G' into the buffer (the first character of the GET request). Yet the output looks as follows: > [init -> test-http_clnt] got IP address 10.0.2.15 > [init -> test-http_clnt] Create new socket ... > [init -> test-http_clnt] Connect to server ... > [init -> test-http_clnt] Send request... > [init -> test-libc_msg_peek] peek: G > [init -> test-libc_msg_peek] peek: E > [init -> test-libc_msg_peek] peek: T > [init -> test-libc_msg_peek] peek: > [init -> test-libc_msg_peek] peek: / (this is just an extract but you get the point) The problem is that this doesn't only account to the recv call but also any further read call will miss the first "peeked" bytes which completely defeats the purpose of MSG_PEEK. According to the lwip_legacy port MSG_PEEK should be supported since 2007 and I'm not sure why it should stay broken such a long time. > 2007-06-30 Frédéric Bernon > * sockets.h, sockets.c: Implement MSG_PEEK flag for recv/recvfrom functions. As far as I can see I have used it correctly so I'm not sure why it isn't working. You can view the code at [1] and run the test with run/libc_msg_peek. Regards, Johannes [1]: https://github.com/jklmnn/genode/commit/b631a3068024748cc8443d1db4e30b35aae…

2 7

Nested threads
by Johannes Kliemann 19 Jun '18

19 Jun '18

Hi all, are nested threads (e.g. a thread class that owns an object of another thread class and starts it from within its entry() method) a special case? I noticed that the terminal session blocked from the inner thread (write was called from the thread but the write implementation of the terminal session was never executed). Putting the inner thread on the same level as the outer one solved the problem. Regards, Johannes

2 2

out_float
by Steven Harp 19 Jun '18

19 Jun '18

I've noticed that sometimes a double value is rendered strangely in log output. (Genode 18.05, x86_32) E.g. unsigned N1 = 3300; unsigned N2 = 5000 - N1; double div1 = N1 / 1000.0; double div2 = N2 / 1000.0; log(N1,"/1000.=",div1,", ",N2,"/1000. =",div2); // ... [init -> test] 3300/1000.=3.299999, 1700/1000. =1.6a0 I know the log facility is not the place for fancy math-mode, but it would be handy not to have to mentally parse out the extraneous "a". I could file an issue if this is deemed something worth fixing. Maybe there is a workaround short of linking with libc? // Steve

2 1

Memory Page Fault
by Chris Rothrock 18 Jun '18

18 Jun '18

I have been having a problem with Genode/NOVA demo booting using a custom boot module we are developing ever since Genode 17.08 was released and I was hoping to get some help tracing down the problem on the Genode side with this issue. Listed below is the output of two boots of the same build of the demo (in this case 17.11), one without our custom SABLE bootloader, one with. Before 17.08, the hand-off from SABLE to the NOVA kernel was fine, Genode booted and the demo ran as it should have. Now I get PAGE FAULT errors. I also tried downloading the previous version, 17.05 and recreating this build but still get errors, and from what I can tell, this is because the latest NOVA port gets brought in so I believe the problem resides with the way that the NOVA kernel handles modules already in memory at load time of the kernel that is causing this page fault. Are there any thoughts or leads you can assist with in this? As a side note, this SABLE boot loader module loads the Linux kernel flawlessly. Chris Rothrock ########################################################## ### Without SABLE Boot Module ### ########################################################## NOVA Microhypervisor v7-ef4b287 (x86_32): May 10 2018 10:39:11 [gcc 6.3.0] [MBI] [ 0] TSC:3593297 kHz BUS:99813 kHz (measured) [ 0] CORE:0:0:0 15:38:1:0 [0] AMD PRO A10-8750B R7, 12 Compute Cores 4C+8G [ 1] CORE:0:1:0 15:38:1:0 [0] AMD PRO A10-8750B R7, 12 Compute Cores 4C+8G [ 3] CORE:0:3:0 15:38:1:0 [0] AMD PRO A10-8750B R7, 12 Compute Cores 4C+8G [ 2] CORE:0:2:0 15:38:1:0 [0] AMD PRO A10-8750B R7, 12 Compute Cores 4C+8G Hypervisor features SVM Hypervisor reports 4x1 CPUs CPU ID (genode->kernel:package:core:thread) remapping remap (0->0:0:0:0) boot cpu remap (1->1:0:1:0) remap (2->2:0:2:0) remap (3->3:0:3:0) Hypervisor info page contains 27 memory descriptors: core image [00100000,007d4000) binaries region [0022f000,007d4000) free for reuse detected physical memory: 0x00000000 - size: 0x0009e800 use physical memory: 0x00000000 - size: 0x0009e000 detected physical memory: 0x00100000 - size: 0xac722000 use physical memory: 0x00100000 - size: 0xac722000 detected physical memory: 0xadb93000 - size: 0x00001000 use physical memory: 0xadb93000 - size: 0x00001000 detected physical memory: 0xadd9a000 - size: 0x00156000 use physical memory: 0xadd9a000 - size: 0x00156000 detected physical memory: 0xae825000 - size: 0x0004b000 use physical memory: 0xae825000 - size: 0x0004b000 detected physical memory: 0xaeff1000 - size: 0x0000f000 use physical memory: 0xaeff1000 - size: 0x0000f000 detected physical memory: 0x00000000 - size: 0x40000000 ROM modules: ROM: [003a2000,003bad70) acpi_drv ROM: [006f3000,006f4d64) config ROM: [00015000,00016000) core_log ROM: [006f5000,00740dfc) fb_drv ROM: [007a4000,007c5810) global_keys_handler ROM: [00013000,00014000) hypervisor_info_page ROM: [0087e000,008bf238) init ROM: [007c6000,0087d554) launchpad ROM: [00648000,00648253) launchpad.config ROM: [00595000,00647c64) ld.lib.so ROM: [00741000,0078621c) liquid_fb ROM: [00332000,003440c0) nit_focus ROM: [00345000,00363af8) nitlog ROM: [00649000,00692c34) nitpicker ROM: [006a9000,006f23f0) platform_drv ROM: [0000e000,0000f000) platform_info ROM: [003bb000,003dbdb8) pointer ROM: [00364000,003887b0) ps2_drv ROM: [00389000,003a1db8) report_rom ROM: [008c0000,008d6500) rom_filter ROM: [003dc000,0057e8bc) scout ROM: [00787000,007a3fd4) status_bar ROM: [0057f000,00594c08) testnit ROM: [00693000,006a8f68) timer Genode 17.11-221-g8d09d02 <local changes> 2718 MiB RAM and 63829 caps assigned to init [init -> acpi_drv] Found MADT [init -> acpi_drv] MADT IRQ 0 -> GSI 2 flags: 0 [init -> acpi_drv] MADT IRQ 9 -> GSI 9 flags: 15 [init -> acpi_drv] Found MCFG [init -> acpi_drv] MCFG BASE 0xe0000000 seg 0x0 bus 0x0-0xff [init -> platform_drv] 0:14.4 - bridge 1:0.0 [init -> platform_drv] 0:15.0 - bridge 2:0.0 [init -> platform_drv] 0:15.2 - bridge 3:0.0 [init -> fb_drv] Found PCI VGA at 00:01.0 [init -> fb_drv] fb mapped to 0x1000 [init -> ps2_drv] Using keyboard with scan code set 1 (xlate) [init -> ps2_drv] Warning: could not reset mouse (missing ack) [init -> ps2_drv] Error: failed to read from port [init -> ps2_drv] Warning: could not reset mouse (unexpected response) [init -> ps2_drv] Error: failed to read from port [init -> ps2_drv] Warning: could not enable stream [init -> fb_drv] Found: VESA BIOS version 3.0 [init -> fb_drv] OEM: AMD ATOMBIOS [init -> platform_drv] PS2 uses IRQ, vector 0x1 [init -> platform_drv] PS2 uses IRQ, vector 0xc [init -> ps2_drv] Warning: setting of mode indicators failed (0xed) [init -> ps2_drv] Warning: setting of mode indicators failed (0xed) [init -> ps2_drv] Warning: setting of mode indicators failed (0xed) [init -> fb_drv] Found: physical frame buffer at 0xc0000000 size: 16777216 [init -> fb_drv] fb mapped to 0xb0000000 [init -> fb_drv] using video mode: 1920x1080@16 [init -> scout] png is 101 x 92, depth=8 ########################################################## ### With SABLE Boot Module ### ########################################################## NOVA Microhypervisor v7-ef4b287 (x86_64): Feb 9 2018 14:02:24 [gcc 6.3.0] [MBI] [ 0] TSC:3593306 kHz BUS:99814 kHz (measured) [ 0] CORE:0:0:0 15:38:1:0 [0] AMD CPU booted by SABLE [ 1] CORE:0:1:0 15:38:1:0 [0] AMD CPU booted by SABLE [ 2] CORE:0:2:0 15:38:1:0 [0] AMD PRO A10-8750B R7, 12 Compute Cores 4C+8G [ 3] CORE:0:3:0 15:38:1:0 [0] AMD PRO A10-8750B R7, 12 Compute Cores 4C+8G Hypervisor features SVM Hypervisor reports 4x1 CPUs CPU ID (genode->kernel:package:core:thread) remapping remap (0->0:0:0:0) boot cpu remap (1->1:0:1:0) remap (2->2:0:2:0) remap (3->3:0:3:0) Hypervisor info page contains 27 memory descriptors: core image [0000000000100000,0000000000902000) binaries region [000000000027a000,0000000000902000) free for reuse detected physical memory: 0x0000000000000000 - size: 0x000000000009e800 use physical memory: 0x0000000000000000 - size: 0x000000000009e000 detected physical memory: 0x0000000000100000 - size: 0x00000000ac722000 use physical memory: 0x0000000000100000 - size: 0x00000000ac722000 detected physical memory: 0x00000000adb93000 - size: 0x0000000000001000 use physical memory: 0x00000000adb93000 - size: 0x0000000000001000 detected physical memory: 0x00000000add9a000 - size: 0x0000000000156000 use physical memory: 0x00000000add9a000 - size: 0x0000000000156000 detected physical memory: 0x00000000ae825000 - size: 0x000000000004b000 use physical memory: 0x00000000ae825000 - size: 0x000000000004b000 detected physical memory: 0x00000000aeff1000 - size: 0x000000000000f000 use physical memory: 0x00000000aeff1000 - size: 0x000000000000f000 detected physical memory: 0x0000000100000000 - size: 0x0000000040000000 use physical memory: 0x0000000100000000 - size: 0x0000000040000000 ROM modules: ROM: [0000000000351000,0000000000370210) acpi_drv ROM: [0000000000719000,000000000071ad61) config ROM: [00000000addb0000,00000000addb1000) core_log ROM: [000000000071b000,0000000000778ed0) fb_drv ROM: [00000000007ed000,00000000008173a8) global_keys_handler ROM: [00000000addae000,00000000addaf000) hypervisor_info_page ROM: [00000000008df000,0000000000933380) init ROM: [0000000000818000,00000000008de1c0) launchpad ROM: [0000000000646000,0000000000646252) launchpad.config ROM: [0000000000565000,0000000000645908) ld.lib.so ROM: [0000000000779000,00000000007c9638) liquid_fb ROM: [00000000002ca000,00000000002dfc70) nit_focus ROM: [00000000002e0000,0000000000302fd0) nitlog ROM: [0000000000647000,00000000006a1500) nitpicker ROM: [00000000006be000,00000000007185d0) platform_drv ROM: [00000000addac000,00000000addad000) platform_info ROM: [0000000000371000,00000000003986e0) pointer ROM: [0000000000303000,0000000000330108) ps2_drv ROM: [0000000000331000,0000000000350198) report_rom ROM: [0000000000934000,00000000009516d8) rom_filter ROM: [0000000000399000,000000000054a9f8) scout ROM: [00000000007ca000,00000000007ec1e0) status_bar ROM: [000000000054b000,0000000000564aa8) testnit ROM: [00000000006a2000,00000000006bdfb8) timer Genode 17.11-221-g8d09d02 <local changes> 3740 MiB RAM and 63253 caps assigned to init Error: PAGE-FAULT IN CORE addr=0xc0 ip=0x1b1950 (read) stack pointer 0xa02fe6f8, qualifiers 0x4 irUwp faulter utcb 0xa02ff000, last message item count 1 0 - type=0x1 rights=0x3 region=0x80000+0x4000 hotspot=0x0(0x801) - delegated #1 0x00000000a02fe6f8 0x00000000001b1950 #2 0x00000000a02fe6f8 0x000000000017ecac -- Thank You, Chris Rothrock Senior System Administrator (315) 308-1637

2 8

depot package version requirements are too strict
by Nobody III 14 Jun '18

14 Jun '18

Currently, if I modify a library or server to fix a bug or add a feature, I can't get other packages to use the updated version without modifying them also. This leads to two problems: 1. unnecessary installation of multiple versions of the same package (unnecessary downloading and disk usage) 2. awkwardness with applying modifications globally For example, if I fix an issue in libc, that fix cannot be cleanly applied to all packages that use libc. It is possible to iterate through all packages that depend on REPO/src/libc/VERSION and modify the version information for each one, but this is awkward and doesn't scale well. To make matters worse, this would have to be (partially) repeated each time a package is downloaded. A potential piece of a solution would be to allow packages to specify the oldest version they are backward-compatible with, e.g. let libc/2018-06-05 specify that it is compatible with libc >= 2018-05-28. For more flexibility, we could also allow packages to specify multiple dependency options, e.g. let noux-system depend on bash or bash-minimal. Also, we need some mechanism to allow cross-repo substitutions, e.g. let blarson/src/libc replace genodelabs/src/libc, or vice versa. For security reasons, there should be restrictions on which repos are allowed for substitutions, so untrusted/src/libc cannot replace genodelabs/src/libc. Any other thoughts or suggestions?

2 1

Official release of Sculpt TC
by Norman Feske 13 Jun '18

13 Jun '18

Hello everyone, I'm happy and proud to announce the official release of the second incarnation of Sculpt OS. Sculpt for The Curious can be downloaded as a ready-to-use disk image: https://genode.org/download/sculpt The downloadable image is a live system, rescue OS, and installer all in one. Compared to earlier versions, it features a graphical user interface for common tasks like connecting to a wireless network. The actual fun happens in the text-based user interface of the "Leitzentrale", which empowers the user to shape the system live, and to inspect the system's state at any time. For detailed usage instructions, please refer to the accompanied documentation: https://genode.org/documentation/articles/sculpt-tc Cherrs Norman -- Dr.-Ing. Norman Feske Genode Labs https://www.genode-labs.com · https://genode.org Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth

1 0

create recipe for ADA library
by Pirmin Duss 13 Jun '18

13 Jun '18

Hello all, I'm trying to create a recipe to build the ADA library as a package archive [1]. Building the library fails with the following error message: Library ada COMPILE a-except_c.o COMPILE gnat_except.o COMPILE s-secsta.o fatal error, run-time library not installed correctly cannot locate file system.ads compilation abandoned /home/pirmin/GapFruit/genode/repos/base/mk/generic.mk:83: recipe for target 's-secsta.o' failed make[6]: *** [s-secsta.o] Error 1 var/libdeps:42: recipe for target 'ada.lib' failed make[5]: *** [ada.lib] Error 2 system.ads is copied in to the source directory of the depot. An other point that doesn't look nice currently is the way used to obtain the location of the port sources [2]. Is there a way to get rid of the "../ports/" part? [1] https://github.com/trimpim/genode/tree/adalib_recipe [2] https://github.com/trimpim/genode/blob/0d40aa7303aca743dc6270f33ecd3e2a3736… best regards, Pirmin

2 2

Re: create recipe for ADA library
by Martin Stein 13 Jun '18

13 Jun '18

Hi Pirmin, Sorry for the delayed answer. El 04/06/18 a las 09:00, Pirmin Duss escribió: > Hi Martin > > On 03.06.2018 12:26, Martin Stein wrote: >> Hi Pirmin, >> >> El 01/06/18 a las 16:11, Pirmin Duss escribió: >>> An other point that doesn't look nice currently is the way used to >>> obtain the location of the port sources [2]. Is there a way to get rid >>> of the "../ports/" part? >> There exists the select_from_ports function that is used widely. Examples: >> >> ARORA_PORT_DIR := $(call select_from_ports,arora) >> >> LIBPNG_DIR := $(call select_from_ports,libpng)/src/lib/libpng > > > select_from_ports seems to return an empty string when I use > PORT_DIR := $(call select_from_ports,gcc)/src/noux-pkg/gcc/gcc > > the copy action fails as it tries to read files from > /src/noux-pkg/gcc/gcc and not out of the port dir. > > Is it possible, that call select_from_ports is only available in the run > scripts? > > best regards, Pirmin > The select_from_port shouldn't depend on the Run mechanism in general. There should also be no problem with using it for noux packages. I assume that there is something missing in your port description that makes select_from_ports fail. You can find a pretty comprehensive description of what is needed for select_from_ports to work with a given port next to its implementation in [1] (see also _lookup_port_hash_file in the same file). Did this help you? Martin [1] base/mk/util.inc

1 0

(no subject)
by Divya Lakshmanan 13 Jun '18

13 Jun '18

Hello, I am working to load Genode on a custom board. For the Sabrelite board, I have seen in the forums that the load address of the Genode image is 0x10800000. I have been successful with loading Genode on the Sabrelite board. Can someone tell me how I can find the similar load address for my custom board? To be able to boot Genode on it. Thank you.

2 1

Error: "No RM attachment, invalid signal context capability"
by Harsha Sharma 12 Jun '18

12 Jun '18

Hello, I get this error with "make run" in my genode and L4 fiasco kernel project. no RM attachment (READ pf_addr=0 pf_ip=102f56c from 2e3000) Warning: invalid signal-context capability Warning: could not resolve pf=0x0 ip=0x102f56c What can be the reason of this kind of error ? What are pf_addr and pf_ip here (I tried looking in genode book but couln't find anything) ? Any help will be appreciated. Thanks. Regards, Harsha Sharma

2 1

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

users