Re: [akaros] replace Linux or FreeBSD scheduler by Akaros

Hi - On 2020-02-01 at 23:06 Joël Krähemann wrote: > Hi all, > > Just want to say hello. I am new to this group. I am the upstream of > GSequencer. > > http://nongnu.org/gsequencer/ > > I have developed a concurrent tree processing function, each parallel > running function is assigned to

[akaros] Ignoring commits with git blame

Hi - Git 2.23 is out. This has support for blame where it can ignoring commits - in particular ones that do major reformatting, like the "8 space tab" change I did a few months back. At the very least, I suggest running: git config --local --replace-all blame.ignorerevsfile

[akaros] fdisk in a VM in a shell script

Hi - This started as a dirt-simple VM app that uses virtio-block and virtio-cons, and only outputs: fdisk -l (not the full fdisk). It's a pretty easy example of making a VM app - and it is less trivial than "date" or "echo". And with some minor changes, we can support a fully interactive fdisk.

[akaros] Dropping a chan in srv via the shell

I think Ron showed me this a long time ago, and I just finally figured it out. My situation: you have a pipe, and a process is listening on one side. You want to send data into it from the shell. Recall a #pipe has two files: data and data1. Writes to one file come out the other. It's

[akaros] Arachne: Core-Aware Thread Management

Cooperative M:N threading system, built on Linux, using cpusets. Haven't read it fully, but I noticed we got mentioned in their rel work. Looks like an after-thought, TBH. https://www.usenix.org/system/files/osdi18-qin.pdf https://github.com/PlatformLab/Arachne Arachne is a new

[akaros] Filesystems in userspace? we're officially a microkernel!

Hi - I've sorted out how to run a filesystem server in userspace, using a VM. The short version is that a Linux VM is just a process, and it has the code to mount a disk with e.g. ext4 on it. The VM guest runs a 9p server, and Akaros can connect to it using virtio-net and NAT. And of course

[akaros] It only took 5 years... (finished the block 'extra data' work)

Hi - Remember that struct block 'extra data' stuff that Drew and I started nearly 5 years ago? Done. The last of the PANIC_EXTRA()s are gone. For most of those years, there were just a couple of corner cases where we'd panic, and we just never ran into them - until I started mounting 9p

[akaros] Various Updates

Hi - Despite the lack of emails, I'm still working on Akaros. Since it's largely been just me, I decided to not spam you all with patches and whatnot. Here are the major things from the last few months: - Updated our paravirt patches to Linux. We're now at 5.1, with an easy one-patch

Re: [akaros] Re: kernel warning in vprintfmt

#syz fix: kprof: use parsecmd() instead of strncmp on user pointers On 2019-05-01 at 19:24 syzbot wrote: > syzbot has found a reproducer for the following crash on: > > HEAD commit:c0f0e9ec printk: check for user pointers in format string .. > git tree: akaros > console output:

Re: [akaros] syzkaller: testing failed: failed to run ["go" "test" "-short" "./..."]: exit status 1

On 2019-03-22 at 9:22 'Dmitry Vyukov' via Akaros wrote: > On Thu, Mar 21, 2019 at 10:37 PM syzbot > wrote: > > > > Hello, > > > > syzbot found the following crash on: > > > > HEAD commit:dce6e62f docs: add custom anchors to headers (2) > > git tree:

Re: [akaros] Re: assertion failed: buf

On 2019-03-06 at 18:12 Dmitry Vyukov wrote: > Woohoo! 3 more down! I was slacking a bit in the last couple of months. =) -- You received this message because you are subscribed to the Google Groups "Akaros" group. To unsubscribe from this group and stop receiving emails from it, send an

Re: [akaros] Re: assertion failed: va

#syz fix: Check safety of user pointer syscall arguments On 2018-10-12 at 19:02 syzbot wrote: > syzbot has found a reproducer for the following crash on: > > HEAD commit:73001cbb86fe alarm: Clean up condition variable usage > git tree: https://github.com/akaros/akaros.git master >

Re: [akaros] Re: assertion failed: buf

#syz fix: Check safety of user pointer syscall arguments On 2018-11-19 at 14:10 syzbot wrote: > syzbot has found a reproducer for the following crash on: > > HEAD commit:f89b6d306593 alarm: Do not allow callbacks to block > git tree: https://github.com/akaros/akaros.git master >

[akaros] A new kernel polling interface (LWN)

https://lwn.net/Articles/743714/ this is about a year old, but interesting in light of akaros's events. short version: it's a method to poll on linux AIO completions, using a ring buffer in shared memory. not sure if it ever got merged - i'm almost always a year behind on my LWN reading. one of

[akaros] Fun with livelock and TCP

Hi - Despite the radio silence and a general lack of time, I'm still working a bit on Akaros. One of the things that popped out of the Go work was that our alarm/timer code had some issues. Notably, the way we were waiting for alarms to complete so we can unset it had some issues. Although

[akaros] Fun with memory leaks and slab tracing

-- View this online at: https://github.com/brho/akaros/compare/fce2e83252b0...cde6e108d028 Barret Rhoden (7): Replace GET_FRAME_START with get_caller_fp() Integrate 'sofar' into sized_allocs

[akaros] Go update

For those curious, I updated the Go port a bit. Initially, it was to fix things that annoyed me about the build process, but ended up fixing a lot of syscall stuff. - Removed all of the open-coded syscalls, particularly in exec_akaros.go. This was the stuff used in StartProcess(). Now we

[akaros] [PATCH] Make SYS_chdir affect other processes

. Unfortunately, our chdir wasn't doing that since we got rid of the VFS. This fixes a Go test that we had always skipped (TestStartProcess). Signed-off-by: Barret Rhoden --- kern/include/ns.h | 4 ++-- kern/src/ns/sysfile.c | 16 +--- kern/src/syscall.c| 18

[akaros] [PATCH 06/10] bash: Handle errors in echo

ure, but don't have errstr yet. Signed-off-by: Barret Rhoden --- .../bash/akaros-patches/4.3.30/0002-echo.patch | 13 - 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/tools/sys-apps/bash/akaros-patches/4.3.30/0002-echo.patch b/tools/sys-apps/bash/akaros-patches/4.3.30/0

[akaros] [PATCH 01/10] Drop the \0 from readstr()

Userspace can handle determining the length by the return value. Some tools (e.g. GNU grep) don't like the \0 and think the contents are binary. Signed-off-by: Barret Rhoden --- kern/src/ns/util.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/kern/src/ns/util.c b/kern

[akaros] [PATCH 03/10] Add GNU grep to sys-apps

This will get built and installed with the other apps. Signed-off-by: Barret Rhoden --- Makefile | 4 ++ tools/sys-apps/grep/.gitignore| 1 + tools/sys-apps/grep/Makefile | 57 +++ .../grep/akaros-patches

[akaros] [PATCH 07/10] bash: Fix errstr

, but it's good enough. Additionally, there's a change we're printing an old errstr, same as with other apps. Regardless, bash's now reports error strings: $ echo garbage > /prof/mpstat bash: echo: write error: Generic Failure, Bad mpstat option (reset|ipi|on|off) Signed-off-by: Barret Rho

[akaros] [PATCH 10/10] Unset CC for apps-install makes

./configure --help). Signed-off-by: Barret Rhoden --- Makefile | 1 + 1 file changed, 1 insertion(+) diff --git a/Makefile b/Makefile index 580139401873..424657bafce4 100644 --- a/Makefile +++ b/Makefile @@ -72,6 +72,7 @@ endef define make_as_parent $(clear_current_env

[akaros] [PATCH 09/10] travis: Dump build logs on failure

Hopefully this will make the travis build failures easier to debug. When/if we add testing, it won't help as much. Maybe we'll need to remove the logs on success, or otherwise communicate that we shouldn't dump the logs. Signed-off-by: Barret Rhoden --- .travis.yml | 9 - 1 file

[akaros] [PATCH 04/10] Print errstr() in glibc's error() (XCC)

% solution. Often when there is no error, it was zeroed, so the only time we'll get an old errstr was when there were multiple errors. Probably. Rebuild glibc. Signed-off-by: Barret Rhoden --- .../glibc-2.19-akaros/sysdeps/akaros/error.c | 339 ++ 1 file changed, 339 insertions

[akaros] [PATCH 00/10] apps-install fixes

d to pull too. Rebuild the world, etc. Barret Rhoden (10): Drop the \0 from readstr() Use readstr() for #device text buffers Add GNU grep to sys-apps Print errstr() in glibc's error() (XCC) bash: Split up and order the akaros-patches bash: Handle errors in echo bash: Fix errstr

[akaros] [PATCH 05/10] bash: Split up and order the akaros-patches

This ensures the patches are applied in order and it is easier to work on various parts of the patch. Signed-off-by: Barret Rhoden --- .../4.3.30/0001-config-etc.patch | 37 .../akaros-patches/4.3.30/0002-echo.patch | 91 + patch => 0003-upstream-scr

[akaros] [PATCH 02/10] Use readstr() for #device text buffers

no concern with running off the end of the sza. Also note that our snprintf() returns the amount written, not the amount of space needed, which is non-standard. Signed-off-by: Barret Rhoden --- kern/drivers/dev/mem.c | 2 +- kern/drivers/dev/proc.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions

Re: [akaros] kernel warning in validstat

#syz dup: kernel warning in statcheck On 2018-09-24 at 21:07 syzbot wrote: > > #syz dup kernel warning in statcheck > > unknown command "dup" > > > > On 2018-07-27 at 15:52 syzbot wrote: > >> Hello, > > >> syzbot found the following crash on: > > >> HEAD commit:8dc899e19d0f

Re: [akaros] kernel warning in syscall

#syz dup: kernel warning in statcheck On 2018-09-24 at 21:07 syzbot wrote: > > #syz dup kernel warning in statcheck > > unknown command "dup" > > > > On 2018-07-31 at 23:02 syzbot wrote: > >> Hello, > > >> syzbot found the following crash on: > > >> HEAD commit:8b21413cae5f

Re: [akaros] kernel warning in sys_fwstat

#syz dup: kernel warning in statcheck On 2018-09-24 at 21:07 syzbot wrote: > > #syz dup kernel warning in statcheck > > unknown command "dup" > > > > On 2018-07-29 at 15:09 syzbot wrote: > >> Hello, > > >> syzbot found the following crash on: > > >> HEAD commit:8dc899e19d0f

Re: [akaros] kernel warning in sysfwstat

#syz dup: kernel warning in statcheck On 2018-09-24 at 21:06 syzbot wrote: > > #syz dup kernel warning in statcheck > > unknown command "dup" > > > > On 2018-08-01 at 7:14 syzbot wrote: > >> Hello, > > >> syzbot found the following crash on: > > >> HEAD commit:8b21413cae5f

Re: [akaros] kernel warning in syswstat

#syz dup: kernel warning in statcheck On 2018-09-24 at 21:06 syzbot wrote: > > #syz dup kernel warning in statcheck > > unknown command "dup" > > > > On 2018-08-02 at 9:25 syzbot wrote: > >> Hello, > > >> syzbot found the following crash on: > > >> HEAD commit:8b21413cae5f

Re: [akaros] kernel warning in validstat

#syz dup kernel warning in statcheck On 2018-07-27 at 15:52 syzbot wrote: > Hello, > > syzbot found the following crash on: > > HEAD commit:8dc899e19d0f vmm: x86: Set the reserved bits in rflags > git tree: https://github.com/akaros/akaros.git/master > console output:

Re: [akaros] kernel warning in syscall

#syz dup kernel warning in statcheck On 2018-07-31 at 23:02 syzbot wrote: > Hello, > > syzbot found the following crash on: > > HEAD commit:8b21413cae5f parlib: Basic implementation of sigaltstack > git tree: https://github.com/akaros/akaros.git/master > console output:

Re: [akaros] kernel warning in sys_fwstat

#syz dup kernel warning in statcheck On 2018-07-29 at 15:09 syzbot wrote: > Hello, > > syzbot found the following crash on: > > HEAD commit:8dc899e19d0f vmm: x86: Set the reserved bits in rflags > git tree: https://github.com/akaros/akaros.git/master > console output:

Re: [akaros] kernel warning in sysfwstat

#syz dup kernel warning in statcheck On 2018-08-01 at 7:14 syzbot wrote: > Hello, > > syzbot found the following crash on: > > HEAD commit:8b21413cae5f parlib: Basic implementation of sigaltstack > git tree: https://github.com/akaros/akaros.git/master > console output:

Re: [akaros] kernel warning in syswstat

#syz dup kernel warning in statcheck On 2018-08-02 at 9:25 syzbot wrote: > Hello, > > syzbot found the following crash on: > > HEAD commit:8b21413cae5f parlib: Basic implementation of sigaltstack > git tree: https://github.com/akaros/akaros.git/master > console output:

[akaros] [PATCH] Throw an error on bad statchecks

statcheck() was returning -1, and both of its callers would throw an error. We might as well report the specific error and squelch the warning. Reported-by: syzbot+0068960e94fbc67ff...@syzkaller.appspotmail.com Signed-off-by: Barret Rhoden --- kern/include/ns.h | 2 +- kern/src/ns

[akaros] [PATCH] Add /tmp to KFS

It is needed by the Go tests. Signed-off-by: Barret Rhoden --- kern/kfs/tmp/.empty | 0 1 file changed, 0 insertions(+), 0 deletions(-) create mode 100644 kern/kfs/tmp/.empty diff --git a/kern/kfs/tmp/.empty b/kern/kfs/tmp/.empty new file mode 100644 index ..e69de29bb2d1

[akaros] [PATCH] Use two-page stacks

xc20598d4>] FP 0xfff75f58 in run_local_syscall 37 [<0xc2059e09>] FP 0xfff75f78 in prep_syscalls 38 [<0xc212206a>] FP 0xfff75f98 in sysenter_callwrapper Signed-off-by: Barret Rhoden --- Kconfig | 12

Re: [akaros] Go 1.4

On 2018-09-07 at 11:18 Dmitry Vyukov wrote: > E.g. for fuchsia we now auto-generate syzkaller descriptions from fidl > service descriptions, so syzkaller knows how to connect to system > services and can talk to them with realistic contents. Sounds pretty cool! -- You received this message

[akaros] Go 1.4

Hi - Thanks to Brian's efforts this summer, we have a working Go 1.4 port that passes all of the Go tests. It's up to date with all the Akaros changes since our 1.3 port a few years ago, to include using SSH and execing test binaries across a 9p mount. Check out the doc for how to use it and

[akaros] [PATCH 07/10] parlib: Support mutex-less condition variables

nt to grab. Signed-off-by: Barret Rhoden --- user/parlib/include/parlib/uthread.h | 8 ++ user/parlib/mutex.c | 93 user/utest/cv.c | 121 +++ 3 files changed, 204 insertions(+), 18 deletions(-) diff --git a/u

[akaros] [PATCH 06/10] parlib: Fix panic output

It would print panic.c:37, instead of the source of the panic. Signed-off-by: Barret Rhoden --- user/parlib/panic.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/user/parlib/panic.c b/user/parlib/panic.c index 44e992e5b17d..f82d5429aec9 100644 --- a/user/parlib/panic.c

[akaros] [PATCH 10/10] Don't cache pcpui in sys_fork()

There wasn't a bug here, yet, but this cuts down on the number of places that could accidentally cache pcpui and access it after a migration. Signed-off-by: Barret Rhoden --- kern/src/syscall.c | 6 ++ 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/kern/src/syscall.c b/kern

[akaros] [PATCH 08/10] parlib: Run alarm handlers outside the tchain lock

ounded to the nearest usec). Signed-off-by: Barret Rhoden --- tests/vmm/vmrunkernel.c| 4 +- user/parlib/alarm.c| 145 - user/parlib/include/parlib/alarm.h | 7 +- user/utest/alarm.c | 4 +- 4 files changed, 87 insertions(

[akaros] [PATCH 04/10] Add dmesg to the monitor

Signed-off-by: Barret Rhoden --- kern/include/kprof.h | 1 + kern/include/monitor.h | 1 + kern/src/monitor.c | 8 3 files changed, 10 insertions(+) diff --git a/kern/include/kprof.h b/kern/include/kprof.h index 3a1150973290..9b8b2aa98d91 100644 --- a/kern/include/kprof.h +++ b

[akaros] [PATCH 05/10] alarm: Do not hold the tchain lock during handlers

-by: Barret Rhoden --- kern/include/alarm.h | 8 +- kern/src/alarm.c | 209 +-- kern/src/rendez.c| 3 +- 3 files changed, 87 insertions(+), 133 deletions(-) diff --git a/kern/include/alarm.h b/kern/include/alarm.h index 1a7e387486eb..127ba1638121

[akaros] [PATCH 03/10] parlib: Fix alarm_abort_sysc() leak

It was possible to set an alarm to abort a syscall and break out of the alarm handler while blocked on an abortable syscall. Signed-off-by: Barret Rhoden --- user/parlib/alarm.c | 23 ++- 1 file changed, 14 insertions(+), 9 deletions(-) diff --git a/user/parlib/alarm.c b

[akaros] [PATCH 02/10] strace: Coalesce a common case statement

SYS_write was doing the same thing that other syscalls were doing; no need for two cases. Signed-off-by: Barret Rhoden --- kern/src/syscall.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/kern/src/syscall.c b/kern/src/syscall.c index eb235385799d..5e89d16ded46 100644 --- a/kern/src

[akaros] [PATCH 00/10] Bughunt

a CV's lock, same as the kernel. This lets you signal and broadcast to CVs from vcore context, and thus most event handlers. It is analogous to calling pthread_cond_signal() from a signal handler - which you aren't allowed to do in POSIX. Barret Rhoden (10): strace: Fix SYS_chdir data parsing

[akaros] LWN - A canary for timer-expiration functions

This is pretty cool for a couple of reasons: https://lwn.net/Articles/731082/ - Exploits bugs to timer code blobs being freed/reused (which Brian can appreciate) - One of the exploits was found with syzkaller. Even for purely debugging reasons, something like this could help our aging alarm

[akaros] Misc fixes, mostly per-cpu stuff

://github.com/brho/akaros/compare/54a1d3c18f8a...e6fa39dfd4e5 Barret Rhoden (9): ktest: Add minor fixes for kernel tests Add support for dynamic per-cpu variables Update per-cpu helpers in compat_todo Update Linux's

[akaros] It's the end of the VFS as we know it

1:01 -0400) View this online at: https://github.com/brho/akaros/compare/4e8396e2d743...54a1d3c18f8a -------- Barret Rhoden (113): Fix 'include' issue slab: Catch attempted NULL frees Add a couple #defines

[akaros] this is cool (updating something previously ported)

So we have a copy of linux's list.h, which I brought over back in 2015. Of course, linux has updated theirs since then, and I'd like the changes. What do we do? Ah! I kept track of the commit: /* Linux's list definitions. * * From commit bc208e0ee0f46744aac95c29366144271a6962bb */ So let's

[akaros] Code review - VMM fixes

Barret Rhoden (5): vmm: Mask NX / unused bits in guest page walks vmm: Rename userspace's gvatogpa() -> gva2gpa() vmm: Handle mov with zero-extend vmm: Add a vmcall for tracing the TF (XCC) vmm: Clean up the VM cmdline files kern/incl

[akaros] Re: Code review - ssh port forwarding

On 2017-12-20 at 13:59 Barret Rhoden <b...@cs.berkeley.edu> wrote: > Easy-peasy. You'll need to rebuild your toolchain and dropbear once I > merge it, pending travis and/or any comments. Merged to master at be1925fb29c9..8c8c78ef9adf (from, to] You can see the entire diff wi

[akaros] Code review - ssh port forwarding

500) View this online at: https://github.com/brho/akaros/compare/be1925fb29c9...8c8c78ef9adf ---- Barret Rhoden (9): net: rock: Keep the ctl FD open (XCC) net: rock: Mirror F_SE

[akaros] [PATCH] mlx4: Linearize large blocks to avoid dropping packets

k is going when you create it. (Though TCP does - we do this with MSS already). Signed-off-by: Barret Rhoden <b...@cs.berkeley.edu> --- Merged to master at 3751ae416e9f..be1925fb29c9 (from, to] You can see the entire diff with 'git diff' or at https://github.com/brho/akaros/comp

Re: [akaros] Code Review - Packed Core allocation strategy

On 2017-12-14 at 00:25 Kevin Klues wrote: > Nice! Valmon‘s legacy lives on. Indeed! Merged to master at ef39a1bbb55a..3751ae416e9f (from, to] You can see the entire diff with 'git diff' or at https://github.com/brho/akaros/compare/ef39a1bbb55a...3751ae416e9f -- You

Re: [akaros] Code Review - Packed Core allocation strategy

17:34:46 -0500) View this online at: https://github.com/brho/akaros/compare/ef39a1bbb55a...3d407708b8bc -------- Barret Rhoden (12): vmx: Squelch per-core startup messages x86: Fix topo

[akaros] Re: Code Review - Fast path for guest IPIs

Merged to master at 4541914c444a..ef39a1bbb55a (from, to] You can see the entire diff with 'git diff' or at https://github.com/brho/akaros/compare/4541914c444a...ef39a1bbb55a On 2017-12-12 at 13:39 Barret Rhoden <b...@cs.berkeley.edu> wrote: > Hi - > > This speeds up IPIs betw

[akaros] [PATCH] Fix Linux timers, jiffies, and HZ

s non-zero, these timers broke. This fixes some nastiness with mod_timer - there never should have been 'delay', since all callers expect to use absolute time. We should have just had jiffies == msec or whatever from the beginning. Signed-off-by: Barret Rhoden <b...@cs.berkeley.edu> --- This fixes the

[akaros] [PATCH] parlib: Fix signature for trace_printf()

t included by glibc's parlib-compat.c, but parlib/stdio.h is. Signed-off-by: Barret Rhoden <b...@cs.berkeley.edu> --- Merged to master at c1257c0d0110..95f45ef968ee (from, to] You can see the entire diff with 'git diff' or at https://github.com/brho/akaros/compare/c1257c0d0110...95f45ef968e

[akaros] Re: Code review - VMM speedups

On 2017-11-22 at 13:39 Barret Rhoden wrote: > vmm: Export the TSC freq via a vmcall (XCC) (2017-11-22 11:49:34 -0500) > > > View this online at: > https://github.com/brho/akaros/compare/770ca98bc691

[akaros] Code review - VMM speedups

Barret Rhoden (9): Remove i386 from glibc (XCC) Add printx to userspace (XCC) x86: Remove UNUSED_ARG #define vmm: Let the VMM control mwait vmexits (XCC) Make cpu_halt() return with IRQs disabled Allow sys_halt_core() to monitor notif_pending vmm: Halt

[akaros] [PATCH] net: tcp: Handle reseqs of length 0

The SACK code was panicking, but it is possible to have length = 0 (on a SYN or FIN). We can just ignore any potential SACK updates in those cases. Signed-off-by: Barret Rhoden <b...@cs.berkeley.edu> --- Merged to master at 734fb3e242e2..770ca98bc691 (from, to] You can see the entir

[akaros] Re: Code review - r8169 and network fixes

On 2017-11-16 at 12:27 Barret Rhoden <b...@cs.berkeley.edu> wrote: > This ports Linux's r8169 driver and fixes a bunch of network and > linux_compat issues along the way. Merged to master at e5d32777b4b9..734fb3e242e2 (from, to] You can see the entire diff with 'git diff'

[akaros] Code review - r8169 and network fixes

) View this online at: https://github.com/brho/akaros/compare/e5d32777b4b9...734fb3e242e2 Barret Rhoden (47): net: tcp: Don't increment snd.nxt net: Move ip.h to net/ip.h net

Re: [akaros] [brho/akaros] kernel panic in generic_file_write (#44)

On 2017-11-08 at 21:58 Kevin Klues wrote: > Has the Go port been updated since I last updated it to 1.3 many moons ago? Wait, I thought you were updating it! No, just kidding. I might get someone to work on it this summer. We'll see. =) Barret -- You received this

Re: [akaros] [brho/akaros] kernel panic in generic_file_write (#44)

On 2017-11-01 at 14:55 Barret Rhoden <b...@cs.berkeley.edu> wrote: > Do you have a log of what the syscall arguments were? Even something > like the last 100 calls would help. Alternatively, I can run the syzkaller. Is there an easy way to seed it or something so I can have it m

Re: [akaros] [brho/akaros] kernel panic in generic_file_write (#44)

On 2017-11-01 at 16:41 Dmitry Vyukov wrote: > Unfortunately our ability to localize and produce reproducers is limited due > to no Go support. Yeah, that makes it a little rough. Do you have a log of what the syscall arguments were? Even something like the last 100

[akaros] Re: Change in akaros[master]: linuxemu: fix fcntl GETFL flags

On 2017-10-26 at 01:16 "Gan Shun Lim (Gerrit)" wrote: > I'd like you to do a code review. Please visit > > https://akaros-review.googlesource.com/4380 > > to review the following change. > > > Change subject: linuxemu: fix fcntl

[akaros] Re: Code review - pagemap bug hunt

On 2017-10-30 at 15:00 Barret Rhoden wrote: > Hi - > > This fixes the bug(s) in https://github.com/brho/akaros/issues/42. > > Also, I expanded the ability to return from panics. Previously, you'd > almost always fail to return from a panic ('exit' or 'e' from the > moni

[akaros] Code review - pagemap bug hunt

Barret Rhoden (4): pm: Fix uninitialized struct page semaphore mm: Don't free pages in the page cache pm: Catch issues with page map pages Add support for attempting returns from panic kern/include/pagemap.h | 1 + kern/include

[akaros] Re: Code review - perf fixes and helpers

On 2017-10-26 at 14:37 Barret Rhoden wrote: > These are bugs and helps I found while fixing up FTQ. The user-visible > changes are helper programs: cpuid (querying cpuid) and turbo (querying > and setting turbo mode). > > You'll need to rebuild perf to get a bug fi

[akaros] Code review - perf fixes and helpers

to 9149ab5e6a29ee449908ed428ede857184108a70: x86: Add get_actual_pstate() (2017-10-26 14:33:29 -0400) View this online at: https://github.com/brho/akaros/compare/0ab512284699...9149ab5e6a29 Barret

Re: [akaros] [brho/akaros] fcntl: reduce noise on invalid arguments (#43)

Thanks! Merged to master at 59d059ae6bad..0ab512284699 (from, to] You can see the entire diff with 'git diff' or at https://github.com/brho/akaros/compare/59d059ae6bad...0ab512284699 On 2017-10-17 at 07:23 Dmitry Vyukov wrote: > Kernel generally should not print to

[akaros] Re: Code review - pth and dup bugs

On 2017-10-16 at 13:58 Barret Rhoden <b...@cs.berkeley.edu> wrote: > The following changes since commit cfdef312b334e97c988db8b524dc712fd0704a4a: > > Add an example jumbo page allocator (2017-10-12 11:23:10 -0700) > > are available in the Git repository at: >

Re: [akaros] Re: [brho/akaros] assertion failed: page && pm_slot_check_refcnt(*page->pg_tree_slot) (#42)

On 2017-10-16 at 18:17 Dmitry Vyukov wrote: > This can be reproduced by running whole fuzzer, though. Thanks, I'll take a look. -- You received this message because you are subscribed to the Google Groups "Akaros" group. To unsubscribe from this group and stop

Re: [akaros] Re: [brho/akaros] assertion failed: page && pm_slot_check_refcnt(*page->pg_tree_slot) (#42)

On 2017-10-16 at 18:09 Dmitry Vyukov wrote: > I failed to create a C reproducer. If I am reading this correctly, sys_exec > is exec system call. Fuzzer itself does not call exec. So I wonder what > calls exec. This probably explains why I can't create a standalone

[akaros] Re: [PATCH] Add an example jumbo page allocator

On 2017-10-12 at 11:24 Barret Rhoden <b...@cs.berkeley.edu> wrote: > For reference. I had one before, but deleted it since it was so simple. > Though it took me longer than I liked to remember how to remake it. Merged to master at 6344ed04e307..cfdef312b334 (from, to] You can se

Re: [akaros] [brho/akaros] assertion failed: page && pm_slot_check_refcnt(*page->pg_tree_slot) (#42)

Hi - On 2017-10-16 at 10:09 Dmitry Vyukov wrote: > I am getting the following crashes. Is it a know issue? If not and you don't > see why it happens right away, I can try to create a reproducer. I don't recognize this one. It looks like there are a couple issues. I'd need to see a little of

Re: [akaros] longjmp from signal handler

Hi - On 2017-10-13 at 11:30 'Dmitry Vyukov' via Akaros wrote: > I need to skip faulting operations in a C program. On linux I do the > following: > [snip] > Is it possible to return from handler and alter thread's context? Just > altering RIP would do, because I can set RIP a another function

[akaros] [PATCH] Add an example jumbo page allocator

For reference. I had one before, but deleted it since it was so simple. Though it took me longer than I liked to remember how to remake it. Signed-off-by: Barret Rhoden <b...@cs.berkeley.edu> --- kern/src/page_alloc.c | 26 ++ 1 file changed, 26 insertions(+) diff

[akaros] Code Review - VMM rebase changes

Barret Rhoden (4): vmm: Attempt to backtrace the guest on error vmm: virtio-net: Handle single IOVs vmm: virtio-blk: Add sanity check vmm: Silently accept another type of PIO out user/vmm/include/vmm/util.h | 4 user/vmm/io.c | 16

[akaros] Re: Code Review - bughunt

Merged to master at 6196e80c8544..416de0097f4f (from, to] You can see the entire diff with 'git diff' or at https://github.com/brho/akaros/compare/6196e80c8544...416de0097f4f Rebuild your toolchain. On 2017-09-29 at 10:54 Barret Rhoden <b...@cs.berkeley.edu> wrote: > Hi - &g

[akaros] Code Review - bughunt

Barret Rhoden (17): Fix sleep x86: Clarify cpuid 64 bit check x86: Update the check for BRK_END vmm: Squelch 'Unsupported IPI' for INIT/SIPI Use AKAROS_ROOT in bt-akaros.sh vfs: Update atime/mtime/ctime

[akaros] Re: Change in akaros[master]: Allow INVPCID from the guest.

On 2017-09-28 at 12:26 "Gan Shun Lim (Gerrit)" wrote: > I'd like you to do a code review. Please visit > > https://akaros-review.googlesource.com/4361 > > to review the following change. > > > Change subject: Allow INVPCID from

[akaros] Re: Code Review - Aggressive FPU save/restore

Merged to master at 88501869ce07..5562ccb05015 (from, to] You can see the entire diff with 'git diff' or at https://github.com/brho/akaros/compare/88501869ce07...5562ccb05015 On 2017-09-27 at 11:14 Barret Rhoden <b...@cs.berkeley.edu> wrote: > Hi - > > This saves and restore

[akaros] Code Review - Aggressive FPU save/restore

...5562ccb05015 Barret Rhoden (3): parlib: Add a couple helpers Add tests for interference parlib: Aggressively save and restore the FPU kern/arch/x86/rdtsc_test.c | 79 tests

[akaros] Re: Change in akaros[master]: Remove deprecated/unused define in linuxemu.h

On 2017-09-22 at 15:28 Gan Shun Lim (Gerrit) wrote: > Gan Shun Lim has posted comments on this change. ( > https://akaros-review.googlesource.com/4320 ) > > Change subject: Remove deprecated/unused define in linuxemu.h > .. >

[akaros] Re: [PATCH] net: tcp: Enforce reasonable 'acked' values

Merged to master at e6f3e2b4f7e1..c2fb441131a4 (from, to] You can see the entire diff with 'git diff' or at https://github.com/brho/akaros/compare/e6f3e2b4f7e1...c2fb441131a4 On 2017-09-19 at 15:17 Barret Rhoden wrote: > If the distant end ACKs something outside the window, specifically an

[akaros] Re: Change in akaros[master]: Added miscellaneous syscalls to linuxemu

On 2017-09-19 at 12:26 "Gan Shun Lim (Gerrit)" wrote: > Gan Shun Lim has posted comments on this change. ( > https://akaros-review.googlesource.com/4300 ) > > Change subject: Added miscellaneous syscalls to linuxemu >

[akaros] [PATCH] net: tcp: Enforce reasonable 'acked' values

be nice to run Akaros on NS or some other simulator and see what happens. Signed-off-by: Barret Rhoden <b...@cs.berkeley.edu> --- kern/src/net/tcp.c | 13 - 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/kern/src/net/tcp.c b/kern/src/net/tcp.c index ba8f35

[akaros] Re: Change in akaros[master]: Implemented additional syscalls in linuxemu

On 2017-09-18 at 17:44 Gan Shun Lim (Gerrit) wrote: > Gan Shun Lim has posted comments on this change. ( > https://akaros-review.googlesource.com/4260 ) > > Change subject: Implemented additional syscalls in linuxemu > .. > >

[akaros] Re: [PATCH 0/2] Removal of some old code

On 2017-09-16 at 11:07 Barret Rhoden wrote: > There are a few chunks of code that we never use and is collecting dust, > notably old virtualization code and the frontend appserver. Merged to master at 738e15028fcd..7d0b77059597 (from, to] You can see the entire diff with 'git diff' or at

[akaros] [PATCH 1/2] Remove unused virtualization code

I looked for any .c file that wasn't in a Kbuild, excluding the Linux mlx4 and bnx2x drivers. I also scanned the header files and checked any suspcious ones to see if they were #included and used. Signed-off-by: Barret Rhoden <b...@cs.berkeley.edu> --- .git

[akaros] [PATCH 2/2] Remove the frontend appserver code

The frontend code was for working with the RISC-V and SPARC boards. Whenever we ever resurrect the RISC-V port, we can add support for whatever they are using then. Signed-off-by: Barret Rhoden <b...@cs.berkeley.edu> --- kern/arch/x86/Kbuild | 1 - kern/arch/x86/frontend.c | 12 ---

  1   2   3   4   5   6   7   >