misc/171520: alc network driver + tso + vlan does not work.
>Number: 171520 >Category: misc >Synopsis: alc network driver + tso + vlan does not work. >Confidential: no >Severity: non-critical >Priority: low >Responsible:freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Mon Sep 10 11:10:02 UTC 2012 >Closed-Date: >Last-Modified: >Originator: Nikolau Nevzorov >Release:FreeBSD 9.0-RELEASE >Organization: >Environment: louna# uname -a FreeBSD louna 9.0-RELEASE FreeBSD 9.0-RELEASE #1 r237140: Sun Jun 17 12:20:32 YEKT 2012 niko@louna:/usr/obj/usr/src/sys/LOUNA amd64 >Description: alc network driver + tso + vlan does not work. alc0: port 0x5000-0x507f mem 0x5700-0x5703 irq 16 at device 0.0 on pci1 alc0: 15872 Tx FIFO, 15360 Rx FIFO alc0: Using 1 MSI message(s). miibus0: on alc0 atphy0: PHY 0 on miibus0 atphy0: none, 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto, auto-flow Hello, I have a system using the ALC network card driver with vlans. By default the driver enables TSO4 support on the card. This support does not seem to work in this configuration and causes the OS to generate packets larger then MTU which are sent to the card to be fragmented. This doesn't seem to happen and the packets are dropped. This causes TCP connections to go extremely slowly as many re-transitions occur. ifconfig alc0 -tso fixes the problem. The problem doesn't occur without vlans. And this problem doesn't occur on routed packet's, only on generated on this host. There was thread on this problem on 2010 but without localisation of problem. http://www.freebsd.org/cgi/query-pr.cgi?pr=147985 >How-To-Repeat: Create vlans on alc interface and try to download via vlan and any file transfer protoco file bigger, than 1mb. >Fix: disable TSO on iface >Release-Note: >Audit-Trail: >Unformatted: ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: bin/169773: sh(1): Resizing causes /bin/sh to repeat edit operations
Mark Johnston wrote:
|On Sat, Sep 08, 2012 at 05:44:56PM +0200, Steffen Daode Nurpmeso wrote:
|> |Synopsis: sh(1): Resizing causes /bin/sh to repeat edit operations
|> |
|> |http://www.freebsd.org/cgi/query-pr.cgi?pr=169773
|>
[.]
|> It's a rather quick first diff for editline(3), i have no more
[.]
|
|I took a closer look at the patch... I think the errno handling is
|mostly ok, except for a couple of places in el_gets() where the return
|value of read_char() isn't stored, and the code ends up looking at an
|uninitialized variable. The attached patch is your patch + the fix for
|this.
|
|> I *think* it effectively results in editline(3) behaving the way
|> it is supposed to work (retrying once after a whatever signal,
|> then failing for a second one). Since el_gets() now fails (as it
|> is supposed to), sh(1) will behave wrong in that the current line
[.]
|>
|> It would be better if editline(3) could be configured to simply
|> restart upon EINTR, or to fixate that behaviour (for FreeBSD)?
|> I don't think it is acceptable to loose a line of user content due
|> to a simple resize?
|> So long and ciao,
|
|Maybe we need a new option for el_set() which sets a flag in
|el->el_signal that determines whether various functions return on EINTR.
|libfetch for example has fetchRestartCalls for this purpose. It's not
|really clear to me why anyone would want EINTR as an error and return
|though.
I have implemented a EL_READRESTART option for editline(3), which
seems to be the easiest approach to get around this.
Other options would have been to implement an el_gets_continue(),
which would have restarted editing with the input of the last
state (but what if that ended in a newline?), or to commit suicide
while trying to deal with signals from within sh(1).
I have also tried to extend editline.3 in respect to
EL_UNBUFFERED, which is only partially documented sofar, and the
yet completely undocumented errno handling.
It is a whole series of local commits indeed, but i don't dare to
attach a MBOX or even (horror) send a patch mail-series, and so
i'll simply attach them in order, including the PR bin/170651
patch (laziness). It seems to work. In reversed order:
- 6.diff: Set EL_READRESTART in interactive sh(1) sessions
- 5.diff: Add a new EL_READRESTART option for editline(3)
- 4.diff: Document errno behaviour of el_getc()/el_gets()
- 3.diff: Document EL_UNBUFFERED for el_set()
- 2.diff: Fix editline(3) char read and errno code flow
: This simply reuses your patch.
- 1.diff: Fix PR bin/170651
: (The plain patch. Maybe possible to leave that off.)
|-Mark
Bye and ciao,
--steffen
commit b68195b7d21912bd13b74412db43e2dbecdd4b92
Author: Steffen Daode Nurpmeso
Date: 2012-09-01 17:21:14 +0200
Fix PR bin/170651
diff --git a/bin/sh/histedit.c b/bin/sh/histedit.c
index 6371599..bd47c0d 100644
--- a/bin/sh/histedit.c
+++ b/bin/sh/histedit.c
@@ -67,7 +67,9 @@ __FBSDID("$FreeBSD$");
History *hist; /* history cookie */
EditLine *el; /* editline cookie */
int displayhist;
+int histedit_init;
static FILE *el_in, *el_out, *el_err;
+static int e1v2;
static char *fc_replace(const char *, char *, char *);
static int not_fcnumber(const char *);
@@ -76,12 +78,21 @@ static int str_to_event(const char *, int);
/*
* Set history and editing status. Called whenever the status may
* have changed (figures out what to do).
+ * If force is set then an editline reinit is issued even if the actual edit
+ * mode hasn't changed - necessary after the locale has changed because
+ * editline bases it's decision what is reported or not upon isprint(3)
*/
void
-histedit(void)
+histedit(int force)
{
+ int nedstate;
-#define editing (Eflag || Vflag)
+ if (! histedit_init)
+ return;
+
+ histedit_init = 2;
+ /* options.c ensures these are mutual exclusive */
+ nedstate = (Eflag ? 1 : 0) | (Vflag ? 2 : 0);
if (iflag) {
if (!hist) {
@@ -97,7 +108,7 @@ histedit(void)
else
out2fmt_flush("sh: can't initialize history\n");
}
- if (editing && !el && isatty(0)) { /* && isatty(2) ??? */
+ if (nedstate && ! el && isatty(0)) { /* && isatty(2) ??? */
/*
* turn editing on
*/
@@ -130,17 +141,14 @@ bad:
out2fmt_flush("sh: can't initialize editing\n");
}
INTON;
- } else if (!editing && el) {
+ } else if (! nedstate && el) {
INTOFF;
el_end(el);
el = NULL;
INTON;
}
- if (el) {
- if (Vflag)
- el_set(el, EL_EDITOR, "vi");
- else if (Eflag)
-
Re: bin/169773: sh(1): Resizing causes /bin/sh to repeat edit operations
On Mon, Sep 10, 2012 at 04:56:35PM +0200, Steffen Daode Nurpmeso wrote: > > I have implemented a EL_READRESTART option for editline(3), which > seems to be the easiest approach to get around this. > Other options would have been to implement an el_gets_continue(), > which would have restarted editing with the input of the last > state (but what if that ended in a newline?), or to commit suicide > while trying to deal with signals from within sh(1). > > I have also tried to extend editline.3 in respect to > EL_UNBUFFERED, which is only partially documented sofar, and the > yet completely undocumented errno handling. > > It is a whole series of local commits indeed, but i don't dare to > attach a MBOX or even (horror) send a patch mail-series, and so > i'll simply attach them in order, including the PR bin/170651 > patch (laziness). It seems to work. In reversed order: > > - 6.diff: Set EL_READRESTART in interactive sh(1) sessions > - 5.diff: Add a new EL_READRESTART option for editline(3) > - 4.diff: Document errno behaviour of el_getc()/el_gets() > - 3.diff: Document EL_UNBUFFERED for el_set() > - 2.diff: Fix editline(3) char read and errno code flow > : This simply reuses your patch. > - 1.diff: Fix PR bin/170651 > : (The plain patch. Maybe possible to leave that off.) I didn't test 1.diff, but the rest of the patches apply and work for me. Thanks, -Mark ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
misc/171524: ipmi driver crashes kernel by reboot or shutdown
>Number: 171524 >Category: misc >Synopsis: ipmi driver crashes kernel by reboot or shutdown >Confidential: no >Severity: non-critical >Priority: low >Responsible:freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Mon Sep 10 16:30:03 UTC 2012 >Closed-Date: >Last-Modified: >Originator: Darko Hojnik >Release:9.1-RC1 and 9.0-RELEASE >Organization: Brainbits GMBH >Environment: FreeBSD mightychicken 9.1-RC1 FreeBSD 9.1-RC1 #0: Tue Aug 14 04:25:06 UTC 2012 r...@farrell.cse.buffalo.edu:/usr/obj/usr/src/sys/GENERIC amd64 >Description: Hi there, if the ipmi driver is loaded the system crashes by reboot or shutdown root@mightychicken:/root # reboot Sep 10 18:39:40 mightychicken reboot: rebooted by root Sep 10 18:39:40 mightychicken syslogd: exiting on signal 15 Waiting (max 60 seconds) for system process `vnlru' to stop...done Waiting (max 60 seconds) for system process `bufdaemon' to stop...done Waiting (max 60 seconds) for system process `syncer' to stop... Syncing disks, vnodes remaining...4 4 4 Sleeping thread (tid 100167, pid 16) owns a non-sleepable lock KDB: stack backtrace of thread 100167: #0 0x808f29e6 at mi_switch+0x186 #1 0x8092b7f2 at sleepq_wait+0x42 #2 0x808f3176 at _sleep+0x376 #3 0x8174a227 at ipmi_submit_driver_request+0x97 #4 0x8174a9df at ipmi_set_watchdog+0xaf #5 0x8174ac8f at ipmi_wd_event+0x8f #6 0x807bb76f at kern_do_pat+0x9f #7 0x80983997 at sched_sync+0x1e7 #8 0x808bb69f at fork_exit+0x11f #9 0x80bc320e at fork_trampoline+0xe panic: sleeping thread cpuid = 19 KDB: stack backtrace: #0 0x80920546 at kdb_backtrace+0x66 #1 0x808ea55e at panic+0x1ce #2 0x8092e9c2 at propagate_priority+0x1d2 #3 0x8092f6ee at turnstile_wait+0x1be #4 0x808d89e8 at _mtx_lock_sleep+0xd8 #5 0x8097d1c3 at vn_syncer_add_to_worklist+0x143 #6 0x809812d4 at reassignbuf+0xe4 #7 0x809659d2 at bdirty+0x42 #8 0x80969a42 at bdwrite+0x52 #9 0x80afcac3 at ffs_update+0x2c3 #10 0x80b09a7f at handle_workitem_remove+0x1cf #11 0x80b08f90 at process_worklist_item+0x2b0 #12 0x80b0cd86 at softdep_process_worklist+0x96 #13 0x80b0f397 at softdep_flush+0x197 #14 0x808bb69f at fork_exit+0x11f #15 0x80bc320e at fork_trampoline+0xe Uptime: 28m48s Automatic reboot in 15 seconds - press a key on the console to abort Rebooting... cpu_reset: Restarting BSP cpu_reset_proxy: Stopped CPU 19 The mainboard is an Tyan S7010 with 96 GB memory and two XEON E5645 by searching on google I've found http://freebsd.1045724.n5.nabble.com/stable-9-panic-on-reboot-ipmi-wd-event-td5728441.html Also I've tried to load the drivers from 9.0 with the same resold. >How-To-Repeat: kldload ipmi reboot >Fix: Workaround kldunload ipmi kldunload smbus reboot >Release-Note: >Audit-Trail: >Unformatted: ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: kern/171520: [alc] alc network driver + tso + vlan does not work.
Old Synopsis: alc network driver + tso + vlan does not work. New Synopsis: [alc] alc network driver + tso + vlan does not work. Responsible-Changed-From-To: freebsd-bugs->freebsd-net Responsible-Changed-By: linimon Responsible-Changed-When: Mon Sep 10 16:30:46 UTC 2012 Responsible-Changed-Why: Over to maintainer(s). http://www.freebsd.org/cgi/query-pr.cgi?pr=171520 ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: kern/171524: [ipmi] ipmi driver crashes kernel by reboot or shutdown
Old Synopsis: ipmi driver crashes kernel by reboot or shutdown New Synopsis: [ipmi] ipmi driver crashes kernel by reboot or shutdown Responsible-Changed-From-To: freebsd-bugs->freebsd-net Responsible-Changed-By: linimon Responsible-Changed-When: Mon Sep 10 16:32:46 UTC 2012 Responsible-Changed-Why: Over to maintainer(s). http://www.freebsd.org/cgi/query-pr.cgi?pr=171524 ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: bin/169773: sh(1): Resizing causes /bin/sh to repeat edit operations
Mark Johnston wrote: |On Mon, Sep 10, 2012 at 04:56:35PM +0200, Steffen Daode Nurpmeso wrote: |> |> I have implemented a EL_READRESTART option for editline(3), which [.] |> - 6.diff: Set EL_READRESTART in interactive sh(1) sessions |> - 5.diff: Add a new EL_READRESTART option for editline(3) |> - 4.diff: Document errno behaviour of el_getc()/el_gets() |> - 3.diff: Document EL_UNBUFFERED for el_set() |> - 2.diff: Fix editline(3) char read and errno code flow |> : This simply reuses your patch. |> - 1.diff: Fix PR bin/170651 |> : (The plain patch. Maybe possible to leave that off.) | |I didn't test 1.diff, but the rest of the patches apply and work for me. Good. |Thanks, |-Mark My pleasure. --steffen ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
misc/171529: Serial chapter introduction mentions obsolete sio(4) interface
>Number: 171529 >Category: misc >Synopsis: Serial chapter introduction mentions obsolete sio(4) interface >Confidential: no >Severity: non-critical >Priority: low >Responsible:freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Mon Sep 10 23:30:02 UTC 2012 >Closed-Date: >Last-Modified: >Originator: Garrett Cooper >Release:9.1-PRERELEASE >Organization: EMC Isilon >Environment: FreeBSD wf048.west.isilon.com 9.1-RC1 FreeBSD 9.1-RC1 #0: Fri Aug 17 02:18:40 PDT 2012 r...@wf048.west.isilon.com:/usr/obj/mnt/freebsd/releng/9.1/sys/ISI-GENERIC amd64 >Description: The link to the following page in the handbook mentions sio(4) instead of uart(4): http://www.freebsd.org/doc/en/books/handbook/serial.html#SERIAL-CABLES-PORTS . The documentation needs to be updated for FreeBSD 8.x+ as sio(4) is defunct in 8.x+: wf048# dmesg | grep ^sio wf048# dmesg | grep ^uart uart0: <16550 or compatible> port 0x3f8-0x3ff irq 4 flags 0x10 on acpi0 uart0: console (115200,n,8,1) uart1: <16550 or compatible> port 0x2f8-0x2ff irq 3 flags 0x10 on acpi0 uart2: <16550 or compatible> port 0x3e8-0x3ef irq 5 flags 0x10 on acpi0 >How-To-Repeat: >Fix: >Release-Note: >Audit-Trail: >Unformatted: ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: docs/171529: Serial chapter introduction mentions obsolete sio(4) interface
Synopsis: Serial chapter introduction mentions obsolete sio(4) interface Class-Changed-From-To: sw-bug->doc-bug Class-Changed-By: gjb Class-Changed-When: Tue Sep 11 01:13:40 UTC 2012 Class-Changed-Why: Docs PR. Responsible-Changed-From-To: freebsd-bugs->freebsd-docs Responsible-Changed-By: gjb Responsible-Changed-When: Tue Sep 11 01:13:40 UTC 2012 Responsible-Changed-Why: Docs PR. http://www.freebsd.org/cgi/query-pr.cgi?pr=171529 ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
kern/171532: ndis(4) driver includes 'pccard'-specific code, even if 'device pccard' absent from config
>Number: 171532 >Category: kern >Synopsis: ndis(4) driver includes 'pccard'-specific code, even if >'device pccard' absent from config >Confidential: no >Severity: non-critical >Priority: low >Responsible:freebsd-bugs >State: open >Quarter: >Keywords: >Date-Required: >Class: sw-bug >Submitter-Id: current-users >Arrival-Date: Tue Sep 11 01:30:03 UTC 2012 >Closed-Date: >Last-Modified: >Originator: Robert Bonomi >Release:8.3-RELEASE i386 >Organization: >Environment: FreeBSD 8.3-RELEAE FreeBSD 8.3-RELEASE #5 i386 >Description: Two 'if' blocks, one at lines 562-569, and the other at lines 1083-1084, in sys/dev/if_ndis/if_ndis.c are RUN-TIME tests for PCMCIA/PC-Card/CARDBUS hardware, and reference routines that exist only in the 'device pccard' support modules. Kernel linking fails if 'device pccard' is not included, even if the machine has no pccard hardware. >How-To-Repeat: attempt to compile a kernel with option NDISAPI device ndis device wlan and WITHOUT device pcard observe the linker failure. >Fix: bracket each of the 'if ()' blocks identified above with *COMPILE-TIME* '#if' or '#ifdef' conditionals on 'something' that is set only if 'device pccard' is in the config file. >Release-Note: >Audit-Trail: >Unformatted: ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: kern/135485: [modules] [patch] During a shutdown, kernel modules should be unloaded in reverse order in which they where loaded
Synopsis: [modules] [patch] During a shutdown, kernel modules should be unloaded in reverse order in which they where loaded Responsible-Changed-From-To: eadler->freebsd-bugs Responsible-Changed-By: eadler Responsible-Changed-When: Tue Sep 11 01:43:53 UTC 2012 Responsible-Changed-Why: I won't be looking at this PR for a while and I need to clear some out of my queue http://www.freebsd.org/cgi/query-pr.cgi?pr=135485 ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: kern/167671: commit references a PR
The following reply was made to PR kern/167671; it has been noted by GNATS. From: dfil...@freebsd.org (dfilter service) To: bug-follo...@freebsd.org Cc: Subject: Re: kern/167671: commit references a PR Date: Tue, 11 Sep 2012 05:58:43 + (UTC) Author: avg Date: Tue Sep 11 05:58:32 2012 New Revision: 240337 URL: http://svn.freebsd.org/changeset/base/240337 Log: MFC r235777: Make dumptid non-static. It is used by libkvm. PR: kern/167671 MFC slacker: harti Modified: stable/9/sys/kern/kern_shutdown.c Directory Properties: stable/9/sys/ (props changed) Modified: stable/9/sys/kern/kern_shutdown.c == --- stable/9/sys/kern/kern_shutdown.c Tue Sep 11 05:04:59 2012 (r240336) +++ stable/9/sys/kern/kern_shutdown.c Tue Sep 11 05:58:32 2012 (r240337) @@ -148,7 +148,7 @@ static struct dumperinfo dumper; /* our /* Context information for dump-debuggers. */ static struct pcb dumppcb;/* Registers. */ -static lwpid_t dumptid; /* Thread ID. */ +lwpid_t dumptid; /* Thread ID. */ static void poweroff_wait(void *, int); static void shutdown_halt(void *junk, int howto); ___ svn-src-...@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org" ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: kern/167671: [libkvm] [patch] libkvm doesn't initialize vnet
Synopsis: [libkvm] [patch] libkvm doesn't initialize vnet State-Changed-From-To: open->closed State-Changed-By: avg State-Changed-When: Tue Sep 11 06:00:23 UTC 2012 State-Changed-Why: fixed in all supported branches http://www.freebsd.org/cgi/query-pr.cgi?pr=167671 ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
Re: kern/167671: commit references a PR
The following reply was made to PR kern/167671; it has been noted by GNATS. From: dfil...@freebsd.org (dfilter service) To: bug-follo...@freebsd.org Cc: Subject: Re: kern/167671: commit references a PR Date: Tue, 11 Sep 2012 05:59:32 + (UTC) Author: avg Date: Tue Sep 11 05:59:19 2012 New Revision: 240338 URL: http://svn.freebsd.org/changeset/base/240338 Log: MFC r235777: Make dumptid non-static. It is used by libkvm. PR: kern/167671 MFC slacker: harti Modified: stable/8/sys/kern/kern_shutdown.c Directory Properties: stable/8/sys/ (props changed) stable/8/sys/kern/ (props changed) Modified: stable/8/sys/kern/kern_shutdown.c == --- stable/8/sys/kern/kern_shutdown.c Tue Sep 11 05:58:32 2012 (r240337) +++ stable/8/sys/kern/kern_shutdown.c Tue Sep 11 05:59:19 2012 (r240338) @@ -146,7 +146,7 @@ static struct dumperinfo dumper; /* our /* Context information for dump-debuggers. */ static struct pcb dumppcb;/* Registers. */ -static lwpid_t dumptid; /* Thread ID. */ +lwpid_t dumptid; /* Thread ID. */ static void boot(int) __dead2; static void poweroff_wait(void *, int); ___ svn-src-...@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/svn-src-all To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org" ___ freebsd-bugs@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-bugs To unsubscribe, send any mail to "freebsd-bugs-unsubscr...@freebsd.org"
