Hello.
I am trying to run one network application. That application uses network (obviously) and storage. I have tested with tar and rump+ram fs, it is ok. But my network fails when I try to use an ahci device. This is the booting log:
spawn qemu-system-x86_64 -no-kvm -cpu core2duo -net nic,model=e1000 -m 512m -net user -redir tcp:8080::80 -nographic -serial mon:stdio -drive id=disk,file=bin/srv.hda,if=none -device ahci,id=ahci -device ide-drive,drive=disk,bus=ahci.0 -boot d -cdrom var/run/netapp.iso
NOVA Microhypervisor v6-1eba7d3 (x86_32): Mar 10 2014 01:25:16 [gcc 4.7.2]
[ 0] CORE:0:0:0 6:f:b:0 [0] Intel(R) Core(TM)2 Duo CPU T7700 @ 2.40GHz Hypervisor does not feature VMX Hypervisor does not feature SVM Hypervisor reports 1x1 CPU - boot CPU is 0 Hypervisor info page contains 26 memory descriptors: detected physical memory: 0x0000000000000000 - size: 0x9fc00 use physical memory: 0x0000000000000000 - size: 0x9f000 detected physical memory: 0x0000000000100000 - size: 0x1fefe000 use physical memory: 0x0000000000100000 - size: 0x1fefe000 map multi-boot module: physical 0x1f8c3000 -> [0x001c6000-0x0028c2d8) - core map multi-boot module: physical 0x1f98a000 -> [0x00002000-0x0000257d) - config map multi-boot module: physical 0x1f98b000 -> [0x00003000-0x0006e3b0) - init map multi-boot module: physical 0x1f9f7000 -> [0x0006f000-0x000cdbfc) - timer map multi-boot module: physical 0x1fa56000 -> [0xb0000000-0xb0097124) - ld.lib.so map multi-boot module: physical 0x1faee000 -> [0xbff04000-0xbffb4208) - libc.lib.so map multi-boot module: physical 0x1fb9f000 -> [0x000ce000-0x000f26d8) - libm.lib.so map multi-boot module: physical 0x1fbc4000 -> [0x000f3000-0x000f91c4) - pthread.lib.so map multi-boot module: physical 0x1fbcb000 -> [0x000fa000-0x000fc7dc) - libc_log.lib.so map multi-boot module: physical 0x1fbce000 -> [0xb0098000-0xb0185750) - lxip.lib.so map multi-boot module: physical 0x1fcbc000 -> [0xbffb5000-0xbffe3c00) - libc_resolv.lib.so map multi-boot module: physical 0x1fceb000 -> [0xb0186000-0xb02096e8) - netapp map multi-boot module: physical 0x1fd6f000 -> [0xb020a000-0xb0292d04) - nic_drv map multi-boot module: physical 0x1fdf8000 -> [0xb0293000-0xb0307914) - acpi_drv map multi-boot module: physical 0x1fe6d000 -> [0xb0308000-0xb0377be0) - pci_drv map multi-boot module: physical 0x1fedd000 -> [0xb0378000-0xb03cde18) - pci_device_pd map multi-boot module: physical 0x1ff33000 -> [0xb03ce000-0xb04369c0) - ahci map multi-boot module: physical 0x1ff9c000 -> [0xb0437000-0xb0480234) - libc_fuse_ext2.lib.so map multi-boot module: physical 0x1ffe6000 -> [0xbffe4000-0xbfffbf60) - libc_block.lib.so :virt_alloc: Allocator 185bb4 dump: Block: [00002000,00003000) size=00001000 avail=00000000 max_avail=00000000 Block: [00003000,0006f000) size=0006c000 avail=00000000 max_avail=00000000 Block: [0006f000,000ce000) size=0005f000 avail=00000000 max_avail=00000000 Block: [000ce000,000f3000) size=00025000 avail=00000000 max_avail=00003000 Block: [000f3000,000fa000) size=00007000 avail=00000000 max_avail=00000000 Block: [000fa000,000fd000) size=00003000 avail=00000000 max_avail=00003000 Block: [000fd000,00100000) size=00003000 avail=00003000 max_avail=00003000 Block: [001c6000,0028d000) size=000c7000 avail=00000000 max_avail=9fd73000 Block: [0028d000,a0000000) size=9fd73000 avail=9fd73000 max_avail=9fd73000 Block: [b0000000,b0098000) size=00098000 avail=00000000 max_avail=9fd73000 Block: [b0098000,b0186000) size=000ee000 avail=00000000 max_avail=9fd73000 Block: [b0186000,b020a000) size=00084000 avail=00000000 max_avail=00000000 Block: [b020a000,b0293000) size=00089000 avail=00000000 max_avail=9fd73000 Block: [b0293000,b0308000) size=00075000 avail=00000000 max_avail=00000000 Block: [b0308000,b0378000) size=00070000 avail=00000000 max_avail=00000000 Block: [b0378000,b03ce000) size=00056000 avail=00000000 max_avail=00000000 Block: [b03ce000,b0437000) size=00069000 avail=00000000 max_avail=0fa7e000 Block: [b0437000,b0481000) size=0004a000 avail=00000000 max_avail=0fa7e000 Block: [b0481000,bfeff000) size=0fa7e000 avail=0fa7e000 max_avail=0fa7e000 Block: [bff04000,bffb5000) size=000b1000 avail=00000000 max_avail=0fa7e000 Block: [bffb5000,bffe4000) size=0002f000 avail=00000000 max_avail=00000000 Block: [bffe4000,bfffc000) size=00018000 avail=00000000 max_avail=00001000 Block: [bfffc000,bfffd000) size=00001000 avail=00001000 max_avail=00001000 => mem_size=2951938048 (2815 MB) / mem_avail=2944356352 (2807 MB) :phys_alloc: Allocator 185758 dump: Block: [00001000,0009f000) size=0009e000 avail=0009e000 max_avail=0009e000 Block: [00100000,00400000) size=00300000 avail=00300000 max_avail=1d0c3000 Block: [02800000,1f8c3000) size=1d0c3000 avail=1d0c3000 max_avail=1d0c3000 => mem_size=491130880 (468 MB) / mem_avail=491130880 (468 MB) :io_mem_alloc: Allocator 18a45c dump: Block: [00000000,00001000) size=00001000 avail=00001000 max_avail=00001000 Block: [0009f000,00100000) size=00061000 avail=00061000 max_avail=00061000 Block: [00400000,02800000) size=02400000 avail=02400000 max_avail=e073c000 Block: [1f8c3000,fffff000) size=e073c000 avail=e073c000 max_avail=e073c000 => mem_size=3803832320 (3627 MB) / mem_avail=3803832320 (3627 MB) int main(): --- create local services --- int main(): --- start init --- int main(): transferred 468 MB to init int main(): --- init created, waiting for exit condition --- [init -> acpi] available memory for ACPI 2048 kiB, for PCI_DRV 9684 kiB [init -> nic_drv] --- iPXE NIC driver started --- [init -> ahci] --- AHCI driver started --- [init -> netapp] void init_libc_block(): using the libc_block plugin [init -> netapp] libc_fuse_ext2: try to mount /dev/blkdev... [init -> acpi] void Acpi_table::_parse_tables(T*, Genode::uint32_t) [with T = unsigned int; Genode::uint32_t = unsigned int]: Found MADT [init -> acpi] MADT IRQ 0 -> GSI 2 flags: 0 [init -> acpi] MADT IRQ 5 -> GSI 5 flags: d [init -> acpi] MADT IRQ 9 -> GSI 9 flags: d [init -> acpi] MADT IRQ 10 -> GSI 10 flags: d [init -> acpi] MADT IRQ 11 -> GSI 11 flags: d [init -> acpi -> pci_drv] PCI driver started [init -> acpi -> pci_drv -> pci_device_pd] PCI device pd starting ... [init -> acpi -> pci_drv -> pci_device_pd] PCI device pd started [init -> acpi] ACPI table format not supported - will not rewrite GSIs [init -> ahci] Found AHCI HBA (Vendor ID: 8086 Device ID: 2922 Class: 00010601) [init -> ahci] Port 0: ATAPI no [init -> ahci] Port 0: Detected interface is active [init -> ahci] AHCI Version: 1.0000 [init -> ahci] CAPs: [init -> ahci] Port count: 6 [init -> ahci] Command slots: 32 [init -> ahci] AHCI only: no [init -> ahci] Native command queuing: yes [init -> ahci] 64 Bit: no [init -> ahci] CMD.ST bit set during device reset --> unknown behavior [init -> netapp] fuse* fuse_new(fuse_chan*, fuse_args*, const fuse_operations*, size_t, void*): [init -> netapp] Using DHCP for interface configuration. [init -> netapp] void init_libc_lxip(): init_libc_lxip() address config=dhcp [init -> netapp] INFO: NET: Registered protocol family 16 [init -> netapp] INFO: NET: Registered protocol family 2 [init -> netapp] INFO: Hash tables configured (established 524288 bind 524288) [init -> netapp] INFO: reno registered [init -> netapp] INFO: cubic registered Quota exceeded! amount=8192, size=4096, consumed=8192 [init -> netapp] upgrading quota donation for SIGNAL session [init -> nic_drv] --- init iPXE NIC [init -> nic_drv] scan_pci(): Found: 00:03.0 8086:100e (rev 03) IRQ 0b [init -> nic_drv] probe_pci_device(): using driver 82540em [init -> nic_drv] adjust_pci_device(): PCI BIOS has not enabled device 00:03.0! Updating PCI command 0103->0107 [init -> nic_drv] adjust_pci_device(): PCI device 00:03.0 latency timer is unreasonably low at 0. Setting to 32. [init -> nic_drv] ioremap(): bus_addr = febc0000 len = 20000 [init -> nic_drv] snprintf not implemented [init -> nic_drv] number of devices: 1 [init -> nic_drv] --- init rx_callbacks [init -> nic_drv] --- get MAC address 52:54:00:12:34:56 Quota exceeded! amount=12288, size=4096, consumed=12288 [init -> netapp] upgrading quota donation for SIGNAL session [init -> netapp] Received mac: 52:54:00:12:34:56 [init -> netapp] driver_net_open called
and stop is here.
This is a run script:
set use_ram 0 set use_tar 0 set use_ahci 1
set disk_image "bin/srv.hda" netapp set build_components { core init drivers/timer drivers/pci drivers/nic app/netapp }
set network_driver "nic_drv"
lappend_if [have_spec acpi] build_components drivers/acpi lappend_if [have_spec pci] build_components drivers/pci/device_pd
lappend_if $use_ram build_components server/ram_blk lappend_if $use_ram build_components server/rump_fs
lappend_if $use_tar build_components server/tar_fs
lappend_if $use_ahci build_components drivers/ahci
build $build_components
create_boot_directory
append config { <config> <parent-provides> <service name="ROM"/> <service name="LOG"/> <service name="CAP"/> <service name="RAM"/> <service name="RM"/> <service name="CPU"/> <service name="PD"/> <service name="IRQ"/> <service name="IO_PORT"/> <service name="IO_MEM"/> <service name="SIGNAL"/> </parent-provides> <default-route> <any-service> <parent/> <any-child/> </any-service> </default-route> <start name="timer"> <resource name="RAM" quantum="1M"/> <provides><service name="Timer"/></provides> </start>}
append config { <start name="nic_drv"> <resource name="RAM" quantum="8M"/> <provides><service name="Nic"/></provides> </start>}
append_if [have_spec acpi] config { <start name="acpi"> <resource name="RAM" quantum="12M"/> <binary name="acpi_drv"/> <provides> <service name="PCI"/> <service name="IRQ" /> </provides> <route> <service name="PCI"> <any-child /> </service> <any-service> <parent/> <any-child /> </any-service> </route> </start>}
append_if [expr ![have_spec acpi] && [have_spec pci]] config { <start name="pci_drv"> <resource name="RAM" quantum="6M"/> <provides> <service name="PCI"/> </provides> </start> }
append_if [expr $use_tar] config { <start name="tar_fs"> <resource name="RAM" quantum="10M"/> <provides><service name="File_system"/></provides> <config> <archive name="srv.tar" /> <policy label="netapp" root="/" writeable="yes" /> </config> </start> }
append_if [expr $use_ram] config { <start name="ram_blk"> <resource name="RAM" quantum="64M"/> <provides><service name="Block"/></provides> <config file="ext2.raw" block_size="512"/> </start> <start name="rump_fs"> <resource name="RAM" quantum="8M" /> <provides><service name="File_system"/></provides> <config fs="ext2fs"><policy label="" root="/" writeable="yes"/></config> </start>}
append_if $use_ahci config { <start name="ahci"> <binary name="ahci" /> <resource name="RAM" quantum="10M" /> <provides><service name="Block" /></provides> <route>}
append_if [expr $use_ahci && [have_spec acpi]] config { <service name="IRQ"><child name="acpi" /></service>} append_if $use_ahci config { <any-service> <parent /> <any-child /></any-service> </route> </start>}
append config { <start name="netapp"> <resource name="RAM" quantum="1G" />} append config { <config> <arg value="netapp" /> <libc tx_buf_size="320M" rx_buf_size="320M"} append config {/> </config> </start> }
append config { </config> }
install_config $config
# # Boot modules #
# generic modules set boot_modules { core init timer ld.lib.so libc.lib.so libm.lib.so pthread.lib.so libc_log.lib.so lxip.lib.so libc_resolv.lib.so netapp nic_drv }
# platform-specific modules lappend_if [have_spec acpi] boot_modules acpi_drv lappend_if [have_spec pci] boot_modules pci_drv lappend_if [have_spec nova] boot_modules pci_device_pd
lappend_if $use_ram boot_modules ext2.raw lappend_if $use_ram boot_modules rump.lib.so lappend_if $use_ram boot_modules rump_fs.lib.so lappend_if $use_ram boot_modules ram_blk lappend_if $use_ram boot_modules rump_fs lappend_if $use_ram boot_modules libc_fs.lib.so
lappend_if $use_tar boot_modules tar_fs lappend_if $use_tar boot_modules srv.tar
lappend_if $use_ahci boot_modules ahci lappend_if $use_ahci boot_modules libc_fuse_ext2.lib.so lappend_if $use_ahci boot_modules libc_block.lib.so
build_boot_image $boot_modules
append_if [have_spec x86] qemu_args " -net nic,model=e1000 " append_if [have_spec lan9118] qemu_args " -net nic,model=lan9118 "
append qemu_args " -m 512m -net user -redir tcp:8080::80" append qemu_args " -nographic -serial mon:stdio "
append_if $use_ahci qemu_args " -drive id=disk,file=$disk_image,if=none -device ahci,id=ahci -device ide-drive,drive=disk,bus=ahci.0 -boot d"
run_genode_until forever
So, maybe I forget something in configuration? Something related to IRQ..
Hi Vasily,
So, maybe I forget something in configuration? Something related to IRQ..
I think you are spot on! Your run script routes the IRQ session of the AHCI driver to the ACPI server, which is correct. But for the NIC driver, you are just using the default route, which means that its IRQ session gets routed to the parent (core's IRQ service), which is wrong. You might try adding a route to the NIC driver that routes its IRQ session to the ACPI server as well.
Why then, did the run script work w/o AHCI? Maybe because Qemu uses different IRQs depending on the configurations with and without AHCI.
Btw, are you planning to execute your run script on any platform w/o ACPI support (such as OKL4 or Linux)? If not, you could greatly simplify it by removing all the convoluted conditionals related to '[have_spec acpi]'. This would make it a bit easier to see configuration mistakes.
Cheers Norman
Hello.
So, maybe I forget somethinetappg in configuration? Something related to IRQ..
I think you are spot on! Your run script routes the IRQ session of the AHCI driver to the ACPI server, which is correct. But for the NIC driver, you are just using the default route, which means that its IRQ session gets routed to the parent (core's IRQ service), which is wrong. You might try adding a route to the NIC driver that routes its IRQ session to the ACPI server as well.
Why then, did the run script work w/o AHCI? Maybe because Qemu uses different IRQs depending on the configurations with and without AHCI.
Btw, are you planning to execute your run script on any platform w/o ACPI support (such as OKL4 or Linux)? If not, you could greatly simplify it by removing all the convoluted conditionals related to '[have_spec acpi]'. This would make it a bit easier to see configuration mistakes.
I have removed ACPI and use_tar. This is a new version. Still something related to IRQ does not work. And I have tested without AHCI - use_ram works.
set use_ram 0 set use_ahci 1
set disk_image "bin/srv.hda"
set build_components { core init drivers/timer drivers/nic drivers/pci app/netapp }
set network_driver "nic_drv"
lappend_if [have_spec pci] build_components drivers/pci/device_pd
lappend_if $use_ram build_components server/ram_blk lappend_if $use_ram build_components server/rump_fs
lappend_if $use_ahci build_components drivers/ahci
build $build_components
create_boot_directory
append config { <config> <parent-provides> <service name="ROM"/> <service name="LOG"/> <service name="CAP"/> <service name="RAM"/> <service name="RM"/> <service name="CPU"/> <service name="PD"/> <service name="IRQ"/> <service name="IO_PORT"/> <service name="IO_MEM"/> <service name="SIGNAL"/> </parent-provides> <default-route> <any-service> <parent/> <any-child/> </any-service> </default-route> <start name="timer"> <resource name="RAM" quantum="1M"/> <provides><service name="Timer"/></provides> </start> <start name="pci_drv"> <resource name="RAM" quantum="6M"/> <provides> <service name="PCI"/> </provides> </start>}
append_if [expr $use_ram] config { <start name="ram_blk"> <resource name="RAM" quantum="64M"/> <provides><service name="Block"/></provides> <config file="ext2.raw" block_size="512"/> </start> <start name="rump_fs"> <resource name="RAM" quantum="8M" /> <provides><service name="File_system"/></provides> <config fs="ext2fs"><policy label="" root="/" writeable="yes"/></config> </start>}
append_if $use_ahci config { <start name="ahci"> <binary name="ahci" /> <resource name="RAM" quantum="10M" /> <provides><service name="Block" /></provides> <route> <any-service> <parent /> <any-child /></any-service> </route> </start>}
append config { <start name="nic_drv"> <resource name="RAM" quantum="8M"/> <provides><service name="Nic"/></provides> </start>}
append config { <start name="netapp"> <resource name="RAM" quantum="1G" /> <config> <arg value="netapp" /> <libc tx_buf_size="320M" rx_buf_size="320M"/> </config> </start> }
append config { </config> }
install_config $config
# # Boot modules #
# generic modules set boot_modules { core init timer ld.lib.so libc.lib.so libm.lib.so pthread.lib.so libc_log.lib.so lxip.lib.so libc_resolv.lib.so netapp nic_drv }
# platform-specific modules lappend_if [have_spec pci] boot_modules pci_drv lappend_if [have_spec nova] boot_modules pci_device_pd
lappend_if $use_ram boot_modules ext2.raw lappend_if $use_ram boot_modules rump.lib.so lappend_if $use_ram boot_modules rump_fs.lib.so lappend_if $use_ram boot_modules ram_blk lappend_if $use_ram boot_modules rump_fs lappend_if $use_ram boot_modules libc_fs.lib.so
lappend_if $use_ahci boot_modules ahci lappend_if $use_ahci boot_modules libc_fuse_ext2.lib.so lappend_if $use_ahci boot_modules libc_block.lib.so
build_boot_image $boot_modules
append_if [have_spec x86] qemu_args " -net nic,model=e1000 " append_if [have_spec lan9118] qemu_args " -net nic,model=lan9118 "
append qemu_args " -m 512m -net user -redir tcp:8080::8080" append qemu_args " -nographic -serial mon:stdio "
append_if $use_ahci qemu_args " -drive id=disk,file=$disk_image,if=none -device ahci,id=ahci -device ide-drive,drive=disk,bus=ahci.0 -boot d"
run_genode_until forever
spawn qemu-system-x86_64 -no-kvm -cpu core2duo -net nic,model=e1000 -m 512m -net user -redir tcp:6379::6379 -redir tcp:8080::80 -nographic -serial mon:stdio -drive id=disk,file=bin/srv.hda,if=none -device ahci,id=ahci -device ide-drive,drive=disk,bus=ahci.0 -boot d -cdrom var/run/netapp.iso
NOVA Microhypervisor v6-1eba7d3 (x86_32): Mar 11 2014 16:35:04 [gcc 4.7.2]
[ 0] CORE:0:0:0 6:f:b:0 [0] Intel(R) Core(TM)2 Duo CPU T7700 @ 2.40GHz Hypervisor does not feature VMX Hypervisor does not feature SVM Hypervisor reports 1x1 CPU - boot CPU is 0 Hypervisor info page contains 25 memory descriptors: detected physical memory: 0x0000000000000000 - size: 0x9fc00 use physical memory: 0x0000000000000000 - size: 0x9f000 detected physical memory: 0x0000000000100000 - size: 0x1fefe000 use physical memory: 0x0000000000100000 - size: 0x1fefe000 map multi-boot module: physical 0x1f938000 -> [0x001c6000-0x0028c2d8) - core map multi-boot module: physical 0x1f9ff000 -> [0x00002000-0x00002496) - config map multi-boot module: physical 0x1fa00000 -> [0x00003000-0x0006e3b0) - init map multi-boot module: physical 0x1fa6c000 -> [0x0006f000-0x000cdbfc) - timer map multi-boot module: physical 0x1facb000 -> [0xb0000000-0xb0097124) - ld.lib.so map multi-boot module: physical 0x1fb63000 -> [0xbff04000-0xbffb4208) - libc.lib.so map multi-boot module: physical 0x1fc14000 -> [0x000ce000-0x000f26d8) - libm.lib.so map multi-boot module: physical 0x1fc39000 -> [0x000f3000-0x000f91c4) - pthread.lib.so map multi-boot module: physical 0x1fc40000 -> [0x000fa000-0x000fc7dc) - libc_log.lib.so map multi-boot module: physical 0x1fc43000 -> [0xb0098000-0xb0185750) - lxip.lib.so map multi-boot module: physical 0x1fd31000 -> [0xbffb5000-0xbffe3c00) - libc_resolv.lib.so map multi-boot module: physical 0x1fd60000 -> [0xb0186000-0xb02096e8) - netapp map multi-boot module: physical 0x1fde4000 -> [0xb020a000-0xb0292d04) - nic_drv map multi-boot module: physical 0x1fe6d000 -> [0xb0293000-0xb0302be0) - pci_drv map multi-boot module: physical 0x1fedd000 -> [0xb0303000-0xb0358e18) - pci_device_pd map multi-boot module: physical 0x1ff33000 -> [0xb0359000-0xb03c19c0) - ahci map multi-boot module: physical 0x1ff9c000 -> [0xb03c2000-0xb040b234) - libc_fuse_ext2.lib.so map multi-boot module: physical 0x1ffe6000 -> [0xbffe4000-0xbfffbf60) - libc_block.lib.so :virt_alloc: Allocator 185bb4 dump: Block: [00002000,00003000) size=00001000 avail=00000000 max_avail=00000000 Block: [00003000,0006f000) size=0006c000 avail=00000000 max_avail=00000000 Block: [0006f000,000ce000) size=0005f000 avail=00000000 max_avail=00000000 Block: [000ce000,000f3000) size=00025000 avail=00000000 max_avail=00003000 Block: [000f3000,000fa000) size=00007000 avail=00000000 max_avail=00000000 Block: [000fa000,000fd000) size=00003000 avail=00000000 max_avail=00003000 Block: [000fd000,00100000) size=00003000 avail=00003000 max_avail=00003000 Block: [001c6000,0028d000) size=000c7000 avail=00000000 max_avail=9fd73000 Block: [0028d000,a0000000) size=9fd73000 avail=9fd73000 max_avail=9fd73000 Block: [b0000000,b0098000) size=00098000 avail=00000000 max_avail=9fd73000 Block: [b0098000,b0186000) size=000ee000 avail=00000000 max_avail=9fd73000 Block: [b0186000,b020a000) size=00084000 avail=00000000 max_avail=00000000 Block: [b020a000,b0293000) size=00089000 avail=00000000 max_avail=9fd73000 Block: [b0293000,b0303000) size=00070000 avail=00000000 max_avail=00000000 Block: [b0303000,b0359000) size=00056000 avail=00000000 max_avail=00000000 Block: [b0359000,b03c2000) size=00069000 avail=00000000 max_avail=00000000 Block: [b03c2000,b040c000) size=0004a000 avail=00000000 max_avail=0faf3000 Block: [b040c000,bfeff000) size=0faf3000 avail=0faf3000 max_avail=0faf3000 Block: [bff04000,bffb5000) size=000b1000 avail=00000000 max_avail=0faf3000 Block: [bffb5000,bffe4000) size=0002f000 avail=00000000 max_avail=00000000 Block: [bffe4000,bfffc000) size=00018000 avail=00000000 max_avail=00001000 Block: [bfffc000,bfffd000) size=00001000 avail=00001000 max_avail=00001000 => mem_size=2951938048 (2815 MB) / mem_avail=2944835584 (2808 MB) :phys_alloc: Allocator 185758 dump: Block: [00001000,0009f000) size=0009e000 avail=0009e000 max_avail=0009e000 Block: [00100000,00400000) size=00300000 avail=00300000 max_avail=1d138000 Block: [02800000,1f938000) size=1d138000 avail=1d138000 max_avail=1d138000 => mem_size=491610112 (468 MB) / mem_avail=491610112 (468 MB) :io_mem_alloc: Allocator 18a45c dump: Block: [00000000,00001000) size=00001000 avail=00001000 max_avail=00001000 Block: [0009f000,00100000) size=00061000 avail=00061000 max_avail=00061000 Block: [00400000,02800000) size=02400000 avail=02400000 max_avail=e06c7000 Block: [1f938000,fffff000) size=e06c7000 avail=e06c7000 max_avail=e06c7000 => mem_size=3803353088 (3627 MB) / mem_avail=3803353088 (3627 MB) int main(): --- create local services --- int main(): --- start init --- int main(): transferred 468 MB to init int main(): --- init created, waiting for exit condition --- [init -> pci_drv] PCI driver started [init -> nic_drv] --- iPXE NIC driver started --- [init -> ahci] --- AHCI driver started --- [init -> pci_drv -> pci_device_pd] PCI device pd starting ... [init -> pci_drv -> pci_device_pd] PCI device pd started [init -> netapp] void init_libc_block(): using the libc_block plugin [init -> pci_drv] Could not open file "config" [init -> pci_drv] Could not obtain config file [init -> netapp] libc_fuse_ext2: try to mount /dev/blkdev... [init -> ahci] Found AHCI HBA (Vendor ID: 8086 Device ID: 2922 Class: 00010601) [init -> ahci] Port 0: ATAPI no [init -> ahci] Port 0: Detected interface is active [init -> ahci] AHCI Version: 1.0000 [init -> ahci] CAPs: [init -> ahci] Port count: 6 [init -> ahci] Command slots: 32 [init -> ahci] AHCI only: no [init -> ahci] Native command queuing: yes [init -> ahci] 64 Bit: no [init -> ahci] CMD.ST bit set during device reset --> unknown behavior [init -> netapp] fuse* fuse_new(fuse_chan*, fuse_args*, const fuse_operations*, size_t, void*): [init -> netapp] Using DHCP for interface configuration. [init -> netapp] void init_libc_lxip(): init_libc_lxip() address config=dhcp [init -> netapp] INFO: NET: Registered protocol family 16 [init -> netapp] INFO: NET: Registered protocol family 2 [init -> netapp] INFO: Hash tables configured (established 524288 bind 524288) [init -> netapp] INFO: reno registered [init -> netapp] INFO: cubic registered Quota exceeded! amount=8192, size=4096, consumed=8192 [init -> netapp] upgrading quota donation for SIGNAL session [init -> nic_drv] --- init iPXE NIC [init -> nic_drv] scan_pci(): Found: 00:03.0 8086:100e (rev 03) IRQ 0b [init -> nic_drv] probe_pci_device(): using driver 82540em [init -> nic_drv] adjust_pci_device(): PCI BIOS has not enabled device 00:03.0! Updating PCI command 0103->0107 [init -> nic_drv] adjust_pci_device(): PCI device 00:03.0 latency timer is unreasonably low at 0. Setting to 32. [init -> nic_drv] ioremap(): bus_addr = febc0000 len = 20000 [init -> nic_drv] snprintf not implemented [init -> nic_drv] number of devices: 1 [init -> nic_drv] --- init rx_callbacks [init -> nic_drv] --- get MAC address 52:54:00:12:34:56 Quota exceeded! amount=12288, size=4096, consumed=12288 [init -> netapp] upgrading quota donation for SIGNAL session [init -> netapp] Received mac: 52:54:00:12:34:56 [init -> netapp] driver_net_open called
Hello Vasily,
it seems you misunderstood Norman's remark regarding ACPI...
On Tue, Mar 11, 2014 at 04:44:54PM +0400, Sartakov A. Vasily wrote:
Hello.
So, maybe I forget somethinetappg in configuration? Something related to IRQ..
I think you are spot on! Your run script routes the IRQ session of the AHCI driver to the ACPI server, which is correct. But for the NIC driver, you are just using the default route, which means that its IRQ session gets routed to the parent (core's IRQ service), which is wrong. You might try adding a route to the NIC driver that routes its IRQ session to the ACPI server as well.
Why then, did the run script work w/o AHCI? Maybe because Qemu uses different IRQs depending on the configurations with and without AHCI.
Btw, are you planning to execute your run script on any platform w/o ACPI support (such as OKL4 or Linux)? If not, you could greatly simplify it by removing all the convoluted conditionals related to '[have_spec acpi]'. This would make it a bit easier to see configuration mistakes.
I have removed ACPI and use_tar.
Norman did not propose to remove ACPI support itself but the conditionals controlling its inclusion. If you run Genode on Fiasco-OC or NOVA you definitely need the acpi_drv and appropriate routes for IRQ sessions.
Regards
Hi Vasily,
On 03/11/2014 02:25 PM, Christian Helmuth wrote: Norman did not propose to remove ACPI support itself but the conditionals controlling its inclusion. If you run Genode on Fiasco-OC or NOVA you definitely need the acpi_drv and appropriate routes for IRQ sessions.
Christian is right. With "remove", I was not referring to the ACPI server but the conditionals. You have to keep everything where '[have_spec acpi]' applies and remove everything where '![have_spec acpi]' applies. You did the opposite.
To learn more about the role of the ACPI server, please have a look at:
https://github.com/genodelabs/genode/blob/master/os/src/drivers/acpi/README
Cheers Norman
Yes, you right,
All these things are returned back already.
On 03/11/2014 02:25 PM, Christian Helmuth wrote: Norman did not propose to remove ACPI support itself but the conditionals controlling its inclusion. If you run Genode on Fiasco-OC or NOVA you definitely need the acpi_drv and appropriate routes for IRQ sessions.
Christian is right. With "remove", I was not referring to the ACPI server but the conditionals. You have to keep everything where '[have_spec acpi]' applies and remove everything where '![have_spec acpi]' applies. You did the opposite.
To learn more about the role of the ACPI server, please have a look at:
https://github.com/genodelabs/genode/blob/master/os/src/drivers/acpi/README
Cheers Norman
Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ Genode-main mailing list Genode-main@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/genode-main
So, lets come back to routing of IRQ.
As far I understand, I have to use ACPI for providing PCI and IRQ. right? AHCI and NIC should use IRQ from ACPI, right?
<start name="acpi"> <resource name="RAM" quantum="12M"/> <binary name="acpi_drv"/> <provides> <service name="PCI"/> <service name="IRQ" /> </provides> <route> <service name="PCI"> <any-child /> </service> <any-service> <parent/> <any-child /> </any-service> </route> </start>
<start name="ahci"> <binary name="ahci" /> <resource name="RAM" quantum="10M" /> <provides><service name="Block" /></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>
<start name="nic_drv"> <resource name="RAM" quantum="8M"/> <provides><service name="Nic"/></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>
Should I use «route» section in ACPI? What should I route? PCI?
On 03/11/2014 02:25 PM, Christian Helmuth wrote: Norman did not propose to remove ACPI support itself but the conditionals controlling its inclusion. If you run Genode on Fiasco-OC or NOVA you definitely need the acpi_drv and appropriate routes for IRQ sessions.
Christian is right. With "remove", I was not referring to the ACPI server but the conditionals. You have to keep everything where '[have_spec acpi]' applies and remove everything where '![have_spec acpi]' applies. You did the opposite.
To learn more about the role of the ACPI server, please have a look at:
https://github.com/genodelabs/genode/blob/master/os/src/drivers/acpi/README
Cheers Norman
Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ Genode-main mailing list Genode-main@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/genode-main
-- Sartakov A. Vasily sartakov@...104...
Hi Vasily,
As far I understand, I have to use ACPI for providing PCI and IRQ. right? AHCI and NIC should use IRQ from ACPI, right?
yes.
<start name="acpi"> <resource name="RAM" quantum="12M"/> <binary name="acpi_drv"/> <provides> <service name="PCI"/> <service name="IRQ" /> </provides> <route> <service name="PCI"> <any-child /> </service> <any-service> <parent/> <any-child /> </any-service> </route> </start>
...
Should I use «route» section in ACPI? What should I route? PCI?
You just revealed a superfluous route. There is indeed no point in having the route for PCI specified for the ACPI driver because the ACPI driver will never open a PCI session. Internally, the ACPI driver spawns an instance of the PCI driver as a child process, but this is of no concern for the init configuration.
You can safely remove the "PCI" route. Just make sure that the IRQ session of the ACPI driver is routed to core's IRQ session. The second rule (for '<any-service>') takes care of that because '<parent />' appears before '<any-child />.
We will remove the superfluous route from the README (and the places where this configuration snippet was copied). Thanks for the notice!
Cheers Norman
Ok, let I check again, because nothing happened and I still stuck in QEMU
All of this route all interfaced from core to acpi, and ACPI provides PCI and IRQ: <start name="acpi"> <resource name="RAM" quantum="12M"/> <binary name="acpi_drv"/> <provides> <service name="PCI"/> <service name="IRQ" /> </provides> <route> <any-service> <parent/> <any-child /> </any-service> </route> </start>
Time should use IRQ session from ACPI, and all other (LOG, RM) from core:
<start name="timer"> <resource name="RAM" quantum="1M"/> <provides><service name="Timer"/></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>
AHCI uses all sessions from core instead of IRQ which provided by ACPI:
<start name="ahci"> <binary name="ahci" /> <resource name="RAM" quantum="10M" /> <provides><service name="Block" /></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>
The same thing about nic_drv:
<start name="nic_drv"> <resource name="RAM" quantum="3M"/> <provides><service name="Nic"/></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>
BTW, if I remove AHCI - everything works fine.
You just revealed a superfluous route. There is indeed no point in having the route for PCI specified for the ACPI driver because the ACPI driver will never open a PCI session. Internally, the ACPI driver spawns an instance of the PCI driver as a child process, but this is of no concern for the init configuration.
You can safely remove the "PCI" route. Just make sure that the IRQ session of the ACPI driver is routed to core's IRQ session. The second rule (for '<any-service>') takes care of that because '<parent />' appears before '<any-child />.
We will remove the superfluous route from the README (and the places where this configuration snippet was copied). Thanks for the notice!
Cheers Norman
-- Dr.-Ing. Norman Feske Genode Labs
http://www.genode-labs.com · http://genode.org
Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth
Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ Genode-main mailing list Genode-main@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/genode-main
On Tue, Mar 11, 2014 at 8:52 PM, Sartakov A. Vasily <sartakov@...104...> wrote:
Ok, let I check again, because nothing happened and I still stuck in QEMU
Try using the "q35" machine in qemu which emulates ICH9 intel chipset, it has fixed some ACPI interrupts in the past. Though back then we used Fiasco.OC which only had interrupts working properly on a real machine. https://github.com/Ksys-labs/genode/commit/e32b1f87859db61b5cb4a510056051689...
All of this route all interfaced from core to acpi, and ACPI provides PCI and IRQ: <start name="acpi"> <resource name="RAM" quantum="12M"/> <binary name="acpi_drv"/> <provides> <service name="PCI"/> <service name="IRQ" /> </provides> <route> <any-service> <parent/> <any-child /> </any-service> </route> </start>
Time should use IRQ session from ACPI, and all other (LOG, RM) from core:
<start name="timer"> <resource name="RAM" quantum="1M"/> <provides><service name="Timer"/></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>AHCI uses all sessions from core instead of IRQ which provided by ACPI:
<start name="ahci"> <binary name="ahci" /> <resource name="RAM" quantum="10M" /> <provides><service name="Block" /></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>The same thing about nic_drv:
<start name="nic_drv"> <resource name="RAM" quantum="3M"/> <provides><service name="Nic"/></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>BTW, if I remove AHCI - everything works fine.
You just revealed a superfluous route. There is indeed no point in having the route for PCI specified for the ACPI driver because the ACPI driver will never open a PCI session. Internally, the ACPI driver spawns an instance of the PCI driver as a child process, but this is of no concern for the init configuration.
You can safely remove the "PCI" route. Just make sure that the IRQ session of the ACPI driver is routed to core's IRQ session. The second rule (for '<any-service>') takes care of that because '<parent />' appears before '<any-child />.
We will remove the superfluous route from the README (and the places where this configuration snippet was copied). Thanks for the notice!
Cheers Norman
-- Dr.-Ing. Norman Feske Genode Labs
http://www.genode-labs.com · http://genode.org
Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth
Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ Genode-main mailing list Genode-main@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/genode-main
-- Sartakov A. Vasily sartakov@...104...
Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ Genode-main mailing list Genode-main@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/genode-main
Hello.
As I have found, only VBox emulates ICH9 ok, QEMU/KVM, VmWare are unusable unfortunately. And yes, real hardware is ok.
Norman, is is possible to add 440BX chipset support or something very old like this?
Try using the "q35" machine in qemu which emulates ICH9 intel chipset, it has fixed some ACPI interrupts in the past. Though back then we used Fiasco.OC which only had interrupts working properly on a real machine. https://github.com/Ksys-labs/genode/commit/e32b1f87859db61b5cb4a510056051689...
All of this route all interfaced from core to acpi, and ACPI provides PCI and IRQ: <start name="acpi"> <resource name="RAM" quantum="12M"/> <binary name="acpi_drv"/> <provides> <service name="PCI"/> <service name="IRQ" /> </provides> <route> <any-service> <parent/> <any-child /> </any-service> </route> </start>
Time should use IRQ session from ACPI, and all other (LOG, RM) from core:
<start name="timer"> <resource name="RAM" quantum="1M"/> <provides><service name="Timer"/></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>AHCI uses all sessions from core instead of IRQ which provided by ACPI:
<start name="ahci"> <binary name="ahci" /> <resource name="RAM" quantum="10M" /> <provides><service name="Block" /></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>The same thing about nic_drv:
<start name="nic_drv"> <resource name="RAM" quantum="3M"/> <provides><service name="Nic"/></provides> <route> <service name="IRQ"><child name="acpi" /></service> <any-service> <parent /> <any-child /></any-service> </route> </start>BTW, if I remove AHCI - everything works fine.
You just revealed a superfluous route. There is indeed no point in having the route for PCI specified for the ACPI driver because the ACPI driver will never open a PCI session. Internally, the ACPI driver spawns an instance of the PCI driver as a child process, but this is of no concern for the init configuration.
You can safely remove the "PCI" route. Just make sure that the IRQ session of the ACPI driver is routed to core's IRQ session. The second rule (for '<any-service>') takes care of that because '<parent />' appears before '<any-child />.
We will remove the superfluous route from the README (and the places where this configuration snippet was copied). Thanks for the notice!
Cheers Norman
-- Dr.-Ing. Norman Feske Genode Labs
http://www.genode-labs.com · http://genode.org
Genode Labs GmbH · Amtsgericht Dresden · HRB 28424 · Sitz Dresden Geschäftsführer: Dr.-Ing. Norman Feske, Christian Helmuth
Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ Genode-main mailing list Genode-main@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/genode-main
-- Sartakov A. Vasily sartakov@...104...
Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ Genode-main mailing list Genode-main@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/genode-main
-- Regards, Alexander
Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ Genode-main mailing list Genode-main@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/genode-main
-
Alexander Tarasikov -
Christian Helmuth -
Norman Feske -
Sartakov A. Vasily