gxemul-devel Mailing List for GXemul

Status: Alpha

Brought to you by: gavare

gxemul-devel — Mailing list for discussions regarding GXemul's development

You can subscribe to this list here.

2008	Jan	Feb	Mar (3)	Apr	May	Jun (1)	Jul (2)	Aug	Sep	Oct	Nov	Dec
2009	Jan (2)	Feb	Mar	Apr	May	Jun	Jul	Aug (1)	Sep	Oct	Nov	Dec
2010	Jan (2)	Feb (1)	Mar	Apr	May (2)	Jun	Jul	Aug (3)	Sep	Oct	Nov	Dec
2011	Jan	Feb	Mar	Apr (2)	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
2012	Jan	Feb	Mar	Apr (1)	May	Jun (4)	Jul	Aug	Sep (4)	Oct	Nov	Dec
2013	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct (1)	Nov	Dec
2014	Jan	Feb	Mar	Apr	May	Jun	Jul (1)	Aug	Sep	Oct	Nov	Dec
2015	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct (1)	Nov	Dec
2016	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep (1)	Oct	Nov	Dec
2018	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug (2)	Sep (1)	Oct	Nov (1)	Dec (1)
2019	Jan	Feb	Mar (4)	Apr	May (1)	Jun (14)	Jul (1)	Aug	Sep (1)	Oct	Nov	Dec
2023	Jan	Feb	Mar (7)	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec

Flat | Threaded

1 2 3 > >> (Page 1 of 3)

[GXemul-devel] Booting from HDD? Or: How to start NetBSD without stdin (on a DECstation)?

From: Jan-Benedict G. <jb...@lu...> - 2023-03-30 16:13:53

Attachments: signature.asc

Hi!

I feel a bit silly, but ... I've now integrated GXemul into my scripts
to allow transparently to start either Qemu, GXemul or SIMH, either
with in foreground (with active stdio) or in background (stdin < /dev/null,
stdout+stderr > logfile) for having emulated systems around as
temporary builders.

  It seems to start up well with:

	gxemul -E decstation -e 3max -L tap_1 -l 08:00:2b:29:3b:01 -d R:disk.img


However, with < /dev/null, I just get the config dump and it stops
generating output with "NetBSD/pmax 10.99.2 FFSv1 Primary Bootstrap",
and it seems to try to get input:

(gdb) bt
#0  0x00007fd1f3ea4794 in __GI___select (nfds=nfds@entry=1, readfds=readfds@entry=0x7ffd2dc1a700, writefds=writefds@entry=0x0, 
    exceptfds=exceptfds@entry=0x0, timeout=timeout@entry=0x7ffd2dc1a6f0) at ../sysdeps/unix/sysv/linux/select.c:69
#1  0x000055f709d23151 in d_avail (d=d@entry=0) at ./src/console/console.c:291
#2  0x000055f709d23ae7 in console_stdin_avail (handle=<optimized out>) at ./src/console/console.c:329
#3  console_charavail (handle=<optimized out>) at ./src/console/console.c:357
#4  0x000055f709f1ea4b in lk201_tick (machine=<optimized out>, d=d@entry=0x55f70ac10b20) at ./src/devices/lk201.c:254
#5  0x000055f709ed8607 in dev_dc7085_tick (cpu=0x7fd1ef7ff000, extra=0x55f70ac102c0) at ./src/devices/dev_dc7085.c:153
#6  0x000055f709f2c8ad in machine_run (machine=0x55f70abf21b0) at ./src/machines/machine.c:686
#7  0x000055f709f3d6a1 in emul_run (emul=emul@entry=0x55f70abf1f70) at ./src/core/emul.c:975
#8  0x000055f709d22c36 in main (argc=11, argv=0x7ffd2dc1aae8) at ./src/core/main.c:773

Any chance I can shutdown its stdin? Would be awesome to be able to
start it in background. If nothing works, I'd still try to put it into
a `screen` or the like, but I'd prefer to not wrap it once more...

Thanks,
  Jan-Benedict

--

[GXemul-devel] [PATCH] Update TAP networking to let it work on a Linux host

From: Jan-Benedict G. <jb...@lu...> - 2023-03-28 08:38:59

Attachments: signature.asc

Hi!

There is probably already a (to-be-pushed-to-upstream-SVN) patch
available to enable TAP networking unter Linux, but nonetheless,
here's what I have locally for the time being.

  I'm using it this way (here as an example for a DECstation booting
the NetBSD installer: There are CI builds for NetBSD HEAD running,
storing the ISO and sets (along with the kernel) in tarballs. This is
extracted and GXemul started with it, on a local TAP interface bridged
with a physical wired NIC:

tar xzf /var/cache/laminar/netbsd-rel-pmax-mipsel.tar.gz
dd bs=1M count=1024 if=/dev/zero of=netbsd.img
ISO="$(ls images/*.iso | head -1)"

TAP=threemax
tunctl -t "${TAP}"       || true
ip link set "${TAP}" up  || true
brctl addif br0 "${TAP}" || true


# gxemul @3maxconfig

gxemul  -E decstation -e 3max                           \
        -L threemax                                     \
        -l 10:11:12:13:14:16                            \
        -d netbsd.img                                   \
        -d "6c:${ISO}"                                  \
        -o "-a"                                         \
        -j netbsd pmax/binary/kernel/netbsd-INSTALL.gz


The plan is to prepare as many simulated systems as possible to
exercise as many NetBSD variants as possible. DECstations are just one
example, SIMH is already running (VAX), as well as Qemu (amd64). Seems
that more archs are reachable (alpha with a small Qemu patch, aarch64,
...) These simulators are used to just do test installations (and
throw away the resulting disk images), but my goal is to also install
"master" disk images and run these with overlays. All three simulators
(Qemu, SIMH, GXemul) support this type of configuration, so that's
ideal to start a nice farm of build hosts. The only difference is that
they're getting a fresh overlay and a different MAC address.


  Oh, and here's the Linux TAP patch. There are others around that
also set the IFF_ONE_QUEUE if available (cf.
http://svn.exactcode.de/t2/trunk/package/emulators/gxemul/linux-tap.patch
for example), but I'm not sure if that's needed or really beneficial.

diff --git a/src/net/net_tap.c b/src/net/net_tap.c
index 188276e..d920695 100644
--- a/src/net/net_tap.c
+++ b/src/net/net_tap.c
@@ -48,6 +48,11 @@
 #include "misc.h"
 #include "net.h"
 
+#ifdef __linux__
+#  include <linux/if.h>
+#  include <linux/if_tun.h>
+#endif
+
 /*
  *  net_tap_rx_for_nic():
  *
@@ -166,7 +171,11 @@ bool net_tap_init(struct net *net, const char *tapdev)
 	int fd;
 	int one = 1;
 
+#ifdef __linux__
+	fd = open("/dev/net/tun", O_RDWR);
+#else
 	fd = open(tapdev, O_RDWR);
+#endif
 	if (fd < 0) {
 		debugmsg(SUBSYS_NET, "tap", VERBOSITY_ERROR,
 		    "unable to open tap device '%s': %s",
@@ -174,6 +183,22 @@ bool net_tap_init(struct net *net, const char *tapdev)
 		return false;
 	}
 
+#ifdef __linux__
+	struct ifreq ifr;
+
+	memset(&ifr, 0, sizeof(ifr));
+	ifr.ifr_flags = IFF_TAP | IFF_NO_PI;
+	strncpy(ifr.ifr_name, tapdev, IFNAMSIZ);
+
+	if (ioctl(fd, TUNSETIFF, &ifr) < 0) {
+		debugmsg(SUBSYS_NET, "tap", VERBOSITY_ERROR,
+		    "unable to attach to "
+		    "tap device '%s': %s", tapdev, strerror(errno));
+		close(fd);
+		return false;
+	}
+#endif
+
 	if (ioctl(fd, FIONBIO, &one) < 0) {
 		debugmsg(SUBSYS_NET, "tap", VERBOSITY_ERROR,
 		    "unable to set non-blocking mode on "


MfG, JBG

--

[GXemul-devel] [PATCH] MAC address for emulated hosts

From: Jan-Benedict G. <jb...@lu...> - 2023-03-28 08:20:19

Attachments: signature.asc

Hi!

On Mon, 2023-03-27 22:49:31 +0200, Jan-Benedict Glaw <jb...@lu...> wrote:
> On Sat, 2023-03-25 20:39:37 +0100, Jan-Benedict Glaw <jb...@lu...> wrote:
> >   @Anders: Do you already have private patches to allow for
> > configurable MAC addresses? Or maybe I'd cook up something?
> 
> I had a closer look. One quite simple way to have configurable MAC
> addresses could be to add a command line option (as well as a "mac()"
> option in the config file) to supply a list of comma-separated MAC
> addresses to be stored in struct machine.
> 
>   A helper function could be used to to parse that string and return
> (if available) the n'th MAC address (or an error.) This could be
> called by net_generate_unique_mac() to either pick a configured MAC
> address or place one (as before.)
> 
> @Anders: Would you accept a patch like that? Or have a different idea
> for supplying / parsing MAC addresses?

No docs yet, but use it either with
`-l 10:11:12:13:14:15,10:12:14:16:18:20` on the command line or with
`mac_list("10:11:12:13:14:15")` within a `machine(...)` context inside
the config file. No docs yet. Works for me as far as tested (with just
one NIC on a simulated DECstation, so supplying multiple MAC addresses
wasn't actually tested yet.)

diff --git a/src/core/emul_parse.c b/src/core/emul_parse.c
index a16d80b..6e31465 100644
--- a/src/core/emul_parse.c
+++ b/src/core/emul_parse.c
@@ -224,6 +224,8 @@ static char cur_machine_start_paused[10];
 static char cur_machine_ncpus[10];
 static char cur_machine_n_gfx_cards[10];
 static char cur_machine_serial_nr[10];
+#define MAX_MAC_LIST_LEN	(8*20)
+static char cur_machine_mac_list[MAX_MAC_LIST_LEN];
 static char cur_machine_emulated_hz[10];
 static char cur_machine_memory[10];
 #define	MAX_N_LOAD		15
@@ -357,6 +359,7 @@ static void parse__emul(struct emul *e, FILE *f, int *in_emul, int *line,
 		cur_machine_serial_nr[0] = '\0';
 		cur_machine_emulated_hz[0] = '\0';
 		cur_machine_memory[0] = '\0';
+		cur_machine_mac_list[0] = '\0';
 		return;
 	}
 
@@ -528,6 +531,10 @@ static void parse__machine(struct emul *e, FILE *f, int *in_emul, int *line,
 			e->next_serial_nr = m->serial_nr+1;
 		}
 
+		if (cur_machine_mac_list[0]) {
+			m->mac_list = strdup (cur_machine_mac_list);
+		}
+
 		if (cur_machine_emulated_hz[0]) {
 			m->emulated_hz = mystrtoull(cur_machine_emulated_hz,
 			    NULL, 0);
@@ -617,6 +624,7 @@ static void parse__machine(struct emul *e, FILE *f, int *in_emul, int *line,
 	WORD("force_netboot", cur_machine_force_netboot);
 	WORD("ncpus", cur_machine_ncpus);
 	WORD("serial_nr", cur_machine_serial_nr);
+	WORD("mac_list", cur_machine_mac_list);
 	WORD("n_gfx_cards", cur_machine_n_gfx_cards);
 	WORD("emulated_hz", cur_machine_emulated_hz);
 	WORD("memory", cur_machine_memory);
diff --git a/src/include/machine.h b/src/include/machine.h
index 02ee1f7..67b314b 100644
--- a/src/include/machine.h
+++ b/src/include/machine.h
@@ -119,6 +119,7 @@ struct machine {
 	    is useful when emulating multiple cards in one machine:  */
 	int	serial_nr;
 	int	nr_of_nics;
+	char	*mac_list;
 
 	/*  Main memory:  */
 	struct memory *memory;
diff --git a/src/net/net_misc.c b/src/net/net_misc.c
index f2826b8..fd96559 100644
--- a/src/net/net_misc.c
+++ b/src/net/net_misc.c
@@ -76,6 +76,48 @@ void net_debugaddr(void *addr, int type)
 }
 
 
+static int net_get_configured_mac(struct machine *m, unsigned int n, unsigned char *macbuf)
+{
+	char *start;
+	unsigned int mac[6];
+
+	fprintf (stderr, "Trying to find MAC address %u from \"%s\"\n", n, m->mac_list);
+
+	if (m == NULL || m->mac_list == NULL || macbuf == NULL)
+		return -1;
+
+	/* Find n'th MAC address.  */
+	start = m->mac_list;
+	while (*start != '\0' && n > 0) {
+		if (*start == ',') {
+			n--;
+		}
+		start++;
+	}
+
+	/* Reached end-of-string.  */
+	if (*start == '\0' || n > 0)
+		return -1;
+
+	/* Try to parse MAC address, copy to output buffer.  */
+	fprintf (stderr, "MAC starts with %s\n", start);
+	if (sscanf(start, "%x:%x:%x:%x:%x:%x",
+	           &mac[0], &mac[1], &mac[2],
+	           &mac[3], &mac[4], &mac[5]) == 6) {
+
+		macbuf[0] = mac[0];
+		macbuf[1] = mac[1];
+		macbuf[2] = mac[2];
+		macbuf[3] = mac[3];
+		macbuf[4] = mac[4];
+		macbuf[5] = mac[5];
+
+		return 0;
+	} else {
+		return -1;
+	}
+}
+
 /*
  *  net_generate_unique_mac():
  *
@@ -92,19 +134,23 @@ void net_generate_unique_mac(struct machine *machine, unsigned char *macbuf)
 		return;
 	}
 
-	// Special case: real SGI machines have 0x08 as their first
-	// byte, so let's mimic that.
-	macbuf[0] = machine->machine_type == MACHINE_SGI ? 0x08 : 0x10;
-	macbuf[1] = 0x20;
-	macbuf[2] = 0x30;
-	macbuf[3] = 0;
-	macbuf[4] = 0;
-	/*  NOTE/TODO: This only allows 8 nics per machine!  */
-	macbuf[5] = (machine->serial_nr << 4) + (machine->nr_of_nics << 1);
-
-	if (macbuf[0] & 1 || macbuf[5] & 1) {
-		fatal("Internal error in net_generate_unique_mac().\n");
-		exit(1);
+	/* Use a configured MAC address if available, or generate
+	   something that seems to be reasonable.  */
+	if (net_get_configured_mac (machine, machine->nr_of_nics, macbuf) != 0) {
+		// Special case: real SGI machines have 0x08 as their first
+		// byte, so let's mimic that.
+		macbuf[0] = machine->machine_type == MACHINE_SGI ? 0x08 : 0x10;
+		macbuf[1] = 0x20;
+		macbuf[2] = 0x30;
+		macbuf[3] = 0;
+		macbuf[4] = 0;
+		/*  NOTE/TODO: This only allows 8 nics per machine!  */
+		macbuf[5] = (machine->serial_nr << 4) + (machine->nr_of_nics << 1);
+
+		if (macbuf[0] & 1 || macbuf[5] & 1) {
+			fatal("Internal error in net_generate_unique_mac().\n");
+			exit(1);
+		}
 	}
 
 	/*  TODO: Remember the mac addresses somewhere?  */
diff --git a/src/core/main.c b/src/core/main.c
index f08f68c..0730dba 100644
--- a/src/core/main.c
+++ b/src/core/main.c
@@ -180,6 +180,7 @@ static void usage(bool longusage)
 	    " ISO9660\n            filesystem, -j sets the name of the"
 	    " kernel to load.\n");
 	printf("  -L tapdev enable tap networking using device 'tapdev'\n");
+	printf("  -l mac,,, Comma-separated list of MAC addresses for all network adapters\n");
 	printf("  -M m      emulate m MBs of physical RAM\n");
 	printf("  -N        display nr of instructions/second average, at"
 	    " regular intervals\n");
@@ -267,7 +268,8 @@ static void usage(bool longusage)
  *  Reads command line arguments.
  */
 int get_cmd_args(int argc, char *argv[], struct emul *emul,
-	char ***diskimagesp, int *n_diskimagesp, char **tap_devname)
+	char ***diskimagesp, int *n_diskimagesp, char **tap_devname,
+	char **mac_list)
 {
 	int using_switch_e = 0, using_switch_E = 0;
 	bool using_switch_Z = false;
@@ -280,7 +282,7 @@ int get_cmd_args(int argc, char *argv[], struct emul *emul,
 	struct machine *m = emul_add_machine(emul, NULL);
 
 	const char *opts =
-	    "AC:c:Dd:E:e:GHhI:iJj:k:KL:M:Nn:Oo:p:QqRrSs:TtVvW:"
+	    "AC:c:Dd:E:e:GHhI:iJj:k:KL:l:M:Nn:Oo:p:QqRrSs:TtVvW:"
 #ifdef WITH_X11
 	    "XxY:"
 #endif
@@ -369,6 +371,9 @@ int get_cmd_args(int argc, char *argv[], struct emul *emul,
 		case 'K':
 			debugger_enter_at_end_of_run = true;
 			break;
+		case 'l':
+			*mac_list = strdup(optarg);
+			break;
 		case 'L':
 			*tap_devname = strdup(optarg);
 			break;
@@ -592,6 +597,7 @@ int main(int argc, char *argv[])
 	struct emul *emul;
 	bool using_config_file = false;
 	char *tap_devname = NULL;
+	char *mac_list = NULL;
 	char **diskimages = NULL;
 	int n_diskimages = 0;
 	int i;
@@ -640,7 +646,7 @@ int main(int argc, char *argv[])
 	settings_add(global_settings, "emul", 1,
 	    SETTINGS_TYPE_SUBSETTINGS, 0, emul->settings);
 
-	get_cmd_args(argc, argv, emul, &diskimages, &n_diskimages, &tap_devname);
+	get_cmd_args(argc, argv, emul, &diskimages, &n_diskimages, &tap_devname, &mac_list);
 
 	if (!skip_srandom_call) {
 		struct timeval tv;
@@ -678,6 +684,9 @@ int main(int argc, char *argv[])
 				return 1;
 			}
 
+		/* Add MAC address list to machine.  */
+		emul->machines[0]->mac_list = strdup(mac_list);
+
 		/*  Initialize one emul:  */
 		if (!emul_simple_init(emul, tap_devname)) {
 			fprintf(stderr, "Could not initialize the emulation.\n");


I'd very much appreciate comments on the approach, or hints when
there's a better way.

Thanks,
  Jan-Benedict

--

Re: [GXemul-devel] MAC address for emulated hosts

From: Jan-Benedict G. <jb...@lu...> - 2023-03-27 20:49:44

Attachments: signature.asc

Hi!

On Sat, 2023-03-25 20:39:37 +0100, Jan-Benedict Glaw <jb...@lu...> wrote:
> I just tried to find a way to configure specific MAC addresses for an
> emulated machine. Couldn't find something in the docs and a quick look
> at the sources seems to indicate that, for all emulated NICs,
> net_generate_unique_mac() is called to calculate a MAC address. It
> seems that for two "VMs" run as separate processes, the same MAC would
> be used?
[...]
>   @Anders: Do you already have private patches to allow for
> configurable MAC addresses? Or maybe I'd cook up something?

I had a closer look. One quite simple way to have configurable MAC
addresses could be to add a command line option (as well as a "mac()"
option in the config file) to supply a list of comma-separated MAC
addresses to be stored in struct machine.

  A helper function could be used to to parse that string and return
(if available) the n'th MAC address (or an error.) This could be
called by net_generate_unique_mac() to either pick a configured MAC
address or place one (as before.)

@Anders: Would you accept a patch like that? Or have a different idea
for supplying / parsing MAC addresses?

MfG, JBG

--

Re: [GXemul-devel] MAC address for emulated hosts

From: Jan-Benedict G. <jb...@lu...> - 2023-03-25 20:18:21

Attachments: signature.asc

On Sat, 2023-03-25 20:39:37 +0100, Jan-Benedict Glaw <jb...@lu...> wrote:
> I just tried to find a way to configure specific MAC addresses for an
> emulated machine. Couldn't find something in the docs and a quick look
> at the sources seems to indicate that, for all emulated NICs,
> net_generate_unique_mac() is called to calculate a MAC address. It
> seems that for two "VMs" run as separate processes, the same MAC would
> be used?

net_generate_unique_mac() uses the machine->serial_nr to be encoded
into four bits of the MAC address. That would basically allow for 16
predictable emulated systems. I'd probaby start with that, but in the
end, there'd be good use for more gxemul systems on a network. So the
serial_nr eases my issue a bit, but doesn't solve the general case.

MfG, JBG

--

[GXemul-devel] MAC address for emulated hosts

From: Jan-Benedict G. <jb...@lu...> - 2023-03-25 19:39:51

Attachments: signature.asc

Hi!

I just tried to find a way to configure specific MAC addresses for an
emulated machine. Couldn't find something in the docs and a quick look
at the sources seems to indicate that, for all emulated NICs,
net_generate_unique_mac() is called to calculate a MAC address. It
seems that for two "VMs" run as separate processes, the same MAC would
be used?

  That would not be working well with TAP devices. And the generated
MAC addresses don't work too well for setting up an external DHCP
server. (Some background: I'd like to setup quite a herd of emulated
hosts, based on Qemu, SIMH and gxemul. They'll usually use an overlay
to a static base disk image and the operating systems on the
individual emulated machines will just be different wrt. their MAC
address, thus get different IP addresses. That however implies that I
can actually _supply_ a MAC address.)

  @Anders: Do you already have private patches to allow for
configurable MAC addresses? Or maybe I'd cook up something?

Thanks,
  Jan-Benedict

--

[GXemul-devel] GXemul development activity?

From: Jan-Benedict G. <jb...@lu...> - 2023-03-15 19:07:34

Attachments: signature.asc

Hi!

Not too long ago, I came across GXemul with the intend to use it for
automated testing (NetBSD installation.) As it didn't work right out
of the box running on Linux (issues with TAP devices for proper
networking), I already sent a preliminary patch to the author.

  After having found the SVN repo on Sourceforge, I cloned it locally
(git svn), but realized that the last commit is from about one and a
half year ago. As the HISTORY file suggests more frequent changes, I
suppose I cloned an outdated repository? OTOH, the mailing lists here
seem to be also quite ... silent. I hope GXemul didn't face its end?

Thanks,
  Jan-Benedict
--

[GXemul-devel] Time issues running NetBSD/landisk under gxemul

From: Andreas G. <gs...@gs...> - 2019-09-16 09:02:25

Hi gxemul-devel,

Running the NetBSD test suite on NetBSD-current/landisk under gxemul
0.6.2 on a NetBSD/amd64 host, after a few minutes, time stops
advancing in the virtual machine.

For example, in the console log at

  http://www.gson.org/netbsd/bugs/build/landisk/2019/2019.09.13.08.09.24/test.log

the first hundred or so testcases show a nonzero execution time as
expected:

    file_return_while: [2.409073s] Passed.
    for_break_case: [2.290891s] Passed.
    for_break_compound: [0.336354s] Passed.

but all the ones after that show an execution time of zero:

    for_break_file: [0.000000s] Passed.
    for_break_for: [0.000000s] Passed.
    for_break_func: [0.000000s] Passed.
    for_break_subshell: [0.000000s] Passed.

When the guest has entered this state, a "sleep" command still takes
about the expected amount of real time, but the clock does not advance
during it:

  # date; sleep 10; date
  Mon Sep 16 08:04:47 UTC 2019
  [ten second pause]
  Mon Sep 16 08:04:47 UTC 2019

This does not appear to happen on real hardware, based on the
logs at http://www.netbsd.org/~martin/landisk-atf/ .
-- 
Andreas Gustafsson, gs...@gs...

Re: [GXemul-devel] gxemul serial terminal output stops until key is pressed

From: Chris H. <cmh...@es...> - 2019-07-05 19:25:30

I’ve encountered a similar issue with DECstation 5000/200 (3max) emulation with Ultrix 4.5 on my little GPD Pocket running Ubuntu 19.04 (and preciously 18.10). I’ll see if I can attach with GDB when it happens and figure out what’s going on.

  — Chris

[GXemul-devel] gxemul serial terminal output stops until key is pressed

From: Andreas G. <gs...@gs...> - 2019-06-27 16:29:22

Hi gxemul-devel,

I'm having a strange problem trying to boot NetBSD/hpcmips 8.1 on
gxemul-trunk on Linux. When the install ramdisk kernel boots, the
serial terminal output unexpectedly stops somewhere in the middle of
the boot messages, before showing the sysinst installation menu.

However, if I then press a key, the output continues. Sometimes it
stops again and I have to press a key again multiple times before it
gets all the way to sysinst. This is of course easy to work around
when installing manually, but it makes automated installs using anita
impossible.

I have reproduced this behavior on two different Linux systems:
"gumbo" which runs Debian 9, and an Amazon EC2 t2.micro instance
running Amazon Linux 2 (ami-0b898040803850657).

This problem does not happen when running gxemul on NetBSD 8.0 or
macOS 10.14.5. I have not tested other gxemul versions nor other
guests with respect to this particular problem. gxemul-trunk.tar.gz
was retrieved today and has a sha1 checksum
8bfd6c81e28b0cc182758fc8803490a0e3462af2.

To reproduce the problem, run the following shell script on a Linux
system. If the guest boots all the way to sysinst, exit gxemul
(control-c quit enter) and run the script again; you may need a few
tries, but sooner or later the console output will stop before sysinst
starts.

#!/bin/sh
set -xe
test -f gxemul-trunk.tar.gz || wget http://gavare.se/gxemul/src/gxemul-trunk.tar.gz
test -f netbsd.gz || wget http://ftp.netbsd.org/pub/NetBSD/NetBSD-8.1/hpcmips/installation/netbsd.gz
ls -al gxemul-trunk.tar.gz
shasum gxemul-trunk.tar.gz || true
test -f gxemul-trunk/gxemul || (
gunzip <gxemul-trunk.tar.gz | tar xfv -
cd gxemul-trunk
sh configure
make -j4
)
test -f wd0.img || dd if=/dev/zero of=wd0.img bs=1024k count=1536
gxemul-trunk/gxemul -M 48 -d wd0.img -emobilepro880 netbsd.gz

--
Andreas Gustafsson, gs...@gs...