qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Cornelia Huck <cohuck@redhat.com>
To: peter.maydell@linaro.org
Cc: qemu-devel@nongnu.org, rth@twiddle.net, agraf@suse.de,
	thuth@redhat.com, borntraeger@de.ibm.com, david@redhat.com,
	Halil Pasic <pasic@linux.vnet.ibm.com>,
	Cornelia Huck <cohuck@redhat.com>
Subject: [Qemu-devel] [PULL 19/33] s390x/3270: handle writes of arbitrary length
Date: Fri,  6 Oct 2017 11:59:42 +0200	[thread overview]
Message-ID: <20171006095956.27534-20-cohuck@redhat.com> (raw)
In-Reply-To: <20171006095956.27534-1-cohuck@redhat.com>

From: Halil Pasic <pasic@linux.vnet.ibm.com>

The problem is, that the current implementation places unrealistic and
arbitrary constraints on the length of writes to the device (that is the
outbound requests), by asserting ccw.count being such that that even the
worst case escaped payload will fit an  more or less arbitrary sized
buffer. Actually on protocol level there is nothing to justify such
a limitation.

Another strange thing is the return value which more or less reflects
the size (written) after escaping instead of before escaping. This
is strange, because this return value is used to calculate SCSW.count.

Let us teach 3270 how to deal with arbitrary long writes.

Signed-off-by: Halil Pasic <pasic@linux.vnet.ibm.com>
Acked-by: Christian Borntraeger <borntraeger@de.ibm.com>
Reviewed-by: Dong Jia Shi <bjsdjshi@linux.vnet.ibm.com>
Reported-by: Jason J . Herne <jjherne@linux.vnet.ibm.com>
Tested-by: Jason J . Herne <jjherne@linux.vnet.ibm.com>
Message-Id: <20170920172314.102710-3-pasic@linux.vnet.ibm.com>
Signed-off-by: Cornelia Huck <cohuck@redhat.com>
---
 hw/char/terminal3270.c | 30 ++++++++++++++++++------------
 1 file changed, 18 insertions(+), 12 deletions(-)

diff --git a/hw/char/terminal3270.c b/hw/char/terminal3270.c
index c976a63cc2..a109ce5987 100644
--- a/hw/char/terminal3270.c
+++ b/hw/char/terminal3270.c
@@ -30,7 +30,6 @@ typedef struct Terminal3270 {
     uint8_t inv[INPUT_BUFFER_SIZE];
     uint8_t outv[OUTPUT_BUFFER_SIZE];
     int in_len;
-    int out_len;
     bool handshake_done;
     guint timer_tag;
 } Terminal3270;
@@ -145,7 +144,6 @@ static void chr_event(void *opaque, int event)
 
     /* Ensure the initial status correct, always reset them. */
     t->in_len = 0;
-    t->out_len = 0;
     t->handshake_done = false;
     if (t->timer_tag) {
         g_source_remove(t->timer_tag);
@@ -231,8 +229,9 @@ static int write_payload_3270(EmulatedCcw3270Device *dev, uint8_t cmd)
     Terminal3270 *t = TERMINAL_3270(dev);
     int retval = 0;
     int count = ccw_dstream_avail(get_cds(t));
-
-    assert(count <= (OUTPUT_BUFFER_SIZE - 3) / 2);
+    int bound = (OUTPUT_BUFFER_SIZE - 3) / 2;
+    int len = MIN(count, bound);
+    int out_len = 0;
 
     if (!t->handshake_done) {
         if (!(t->outv[0] == IAC && t->outv[1] != IAC)) {
@@ -247,16 +246,23 @@ static int write_payload_3270(EmulatedCcw3270Device *dev, uint8_t cmd)
         /* We just say we consumed all data if there's no backend. */
         return count;
     }
-    t->outv[0] = cmd;
-    ccw_dstream_read_buf(get_cds(t), &t->outv[1], count);
-    t->out_len = count + 1;
 
-    t->out_len = insert_IAC_escape_char(t->outv, t->out_len);
-    t->outv[t->out_len++] = IAC;
-    t->outv[t->out_len++] = IAC_EOR;
+    t->outv[out_len++] = cmd;
+    do {
+        ccw_dstream_read_buf(get_cds(t), &t->outv[out_len], len);
+        count = ccw_dstream_avail(get_cds(t));
+        out_len += len;
 
-    retval = qemu_chr_fe_write_all(&t->chr, t->outv, t->out_len);
-    return (retval <= 0) ? 0 : (retval - 3);
+        out_len = insert_IAC_escape_char(t->outv, out_len);
+        if (!count) {
+            t->outv[out_len++] = IAC;
+            t->outv[out_len++] = IAC_EOR;
+        }
+        retval = qemu_chr_fe_write_all(&t->chr, t->outv, out_len);
+        len = MIN(count, bound);
+        out_len = 0;
+    } while (len && retval >= 0);
+    return (retval <= 0) ? 0 : get_cds(t)->count;
 }
 
 static Property terminal_properties[] = {
-- 
2.13.6

  parent reply	other threads:[~2017-10-06 10:01 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-10-06  9:59 [Qemu-devel] [PULL 00/33] next batch of s390x patches Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 01/33] s390x/tcg: implement spm (SET PROGRAM MASK) Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 02/33] s390x/tcg: move wrap_address() to internal.h Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 03/33] s390x/tcg: add basic MSA features Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 04/33] s390x: use generic cpu_model parsing Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 05/33] s390x/kvm: fix and cleanup storing CPU status Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 06/33] s390x/css: introduce css data stream Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 07/33] s390x/css: use ccw " Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 08/33] virtio-ccw: " Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 09/33] 390x/css: introduce maximum data address checking Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 10/33] s390x/css: support ccw IDA Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 11/33] s390x/tcg: fix checking for invalid memory check Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 12/33] s390x/tcg: add MMU for real addresses Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 13/33] s390x/tcg: make lura(g) use the new _real mmu Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 14/33] s390x/tcg: make stora(g) " Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 15/33] s390x/tcg: make testblock " Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 16/33] s390x/tcg: make idte/ipte " Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 17/33] Revert "s390x/ccw: create s390 phb conditionally" Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 18/33] s390x/3270: IDA support for 3270 via CcwDataStream Cornelia Huck
2017-10-06  9:59 ` Cornelia Huck [this message]
2017-10-06  9:59 ` [Qemu-devel] [PULL 20/33] MAINTAINERS: use KVM s390x maintainers for kvm-stubs.c and kvm_s390x.h Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 21/33] s390x: raise CPU hotplug irq after really hotplugged Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 22/33] s390x/cpumodel: fix max STFL(E) bit number Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 23/33] target/s390x: get rid of next_core_id Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 24/33] s390x: introduce and use S390_MAX_CPUS Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 25/33] s390x/tcg: make STFL store into the lowcore Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 26/33] s390x: sort some devices into categories Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 27/33] s390x/css: fix css migration compat handling Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 28/33] s390/kvm: Support for get/set of extended TOD-Clock for guest Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 29/33] s390/kvm: make TOD setting failures fatal for migration Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 30/33] s390x/sclp: Mark the sclp device with user_creatable = false Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 31/33] s390x/sclp: mark sclp-cpu-hotplug as non-usercreatable Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 32/33] s390x/tcg: initialize machine check queue Cornelia Huck
2017-10-06  9:59 ` [Qemu-devel] [PULL 33/33] hw/s390x: Mark the "sclpquiesce" device with user_creatable = false Cornelia Huck
2017-10-06 16:00 ` [Qemu-devel] [PULL 00/33] next batch of s390x patches Peter Maydell

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20171006095956.27534-20-cohuck@redhat.com \
    --to=cohuck@redhat.com \
    --cc=agraf@suse.de \
    --cc=borntraeger@de.ibm.com \
    --cc=david@redhat.com \
    --cc=pasic@linux.vnet.ibm.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=rth@twiddle.net \
    --cc=thuth@redhat.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).