qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Sven Schnelle <svens@stackframe.org>
To: Peter Maydell <peter.maydell@linaro.org>
Cc: Paolo Bonzini <pbonzini@redhat.com>,  Fam Zheng <fam@euphon.net>,
	qemu-devel@nongnu.org,  deller@gmx.de
Subject: Re: [PATCH] hw/scsi/lsi53c895a: add hack to prevent scsi timeouts in HP-UX 10.20
Date: Thu, 29 Feb 2024 17:54:09 +0100	[thread overview]
Message-ID: <87r0gvxlce.fsf@t14.stackframe.org> (raw)
In-Reply-To: <CAFEAcA-KioRCJ97dbaiSf0f5uM0=+nAb=oLbFiRg=EPGwd57sw@mail.gmail.com> (Peter Maydell's message of "Thu, 29 Feb 2024 09:36:38 +0000")

Peter Maydell <peter.maydell@linaro.org> writes:

> On Wed, 28 Feb 2024 at 21:12, Sven Schnelle <svens@stackframe.org> wrote:
>>
>> HP-UX 10.20 seems to make the lsi53c895a spinning on a memory location
>> under certain circumstances. As the SCSI controller and CPU are not
>> running at the same time this loop will never finish. After some
>> time, the check loop interrupts with a unexpected device disconnect.
>> This works, but is slow because the kernel resets the scsi controller.
>> Instead of signaling UDC, add an option 'hpux-spin-workaround' which
>> emulates a INTERRUPT 2 script instruction. This instruction tells the
>> kernel that the request was fulfilled. With this change, SCSI speeds
>> improves significantly.
>> [..]
> I see we already have a hacky workaround for other OSes
> that do something similar. The ideal fix for both of these
> I think would be for lsi_execute_script() to, instead of stopping,
> arrange to defer executing more script instructions until
> after the guest has had a chance to run a bit more.
> I think setting a timer that calls lsi_resume_script() after
> a while would have that effect.

Thanks, good idea. So something like this?

diff --git a/hw/scsi/lsi53c895a.c b/hw/scsi/lsi53c895a.c
index d607a5f9fb..9931799d44 100644
--- a/hw/scsi/lsi53c895a.c
+++ b/hw/scsi/lsi53c895a.c
@@ -188,7 +188,7 @@ static const char *names[] = {
 #define LSI_TAG_VALID     (1 << 16)
 
 /* Maximum instructions to process. */
-#define LSI_MAX_INSN    10000
+#define LSI_MAX_INSN    1000
 
 typedef struct lsi_request {
     SCSIRequest *req;
@@ -205,6 +205,7 @@ enum {
     LSI_WAIT_RESELECT, /* Wait Reselect instruction has been issued */
     LSI_DMA_SCRIPTS, /* processing DMA from lsi_execute_script */
     LSI_DMA_IN_PROGRESS, /* DMA operation is in progress */
+    LSI_WAIT_SCRIPTS, /* SCRIPTS where stopped because of instruction count limit */
 };
 
 enum {
@@ -224,6 +225,7 @@ struct LSIState {
     MemoryRegion ram_io;
     MemoryRegion io_io;
     AddressSpace pci_io_as;
+    QEMUTimer *scripts_timer;
 
     int carry; /* ??? Should this be an a visible register somewhere?  */
     int status;
@@ -1152,13 +1154,9 @@ again:
      * which should be enough for all valid use cases).
      */
     if (++insn_processed > LSI_MAX_INSN || reentrancy_level > 8) {
-        if (!(s->sien0 & LSI_SIST0_UDC)) {
-            qemu_log_mask(LOG_GUEST_ERROR,
-                          "lsi_scsi: inf. loop with UDC masked");
-        }
-        lsi_script_scsi_interrupt(s, LSI_SIST0_UDC, 0);
-        lsi_disconnect(s);
-        trace_lsi_execute_script_stop();
+        s->waiting = LSI_WAIT_SCRIPTS;
+        timer_mod(s->scripts_timer, qemu_clock_get_us(QEMU_CLOCK_VIRTUAL) + 500);
+        trace_lsi_scripts_timer_arm();
         reentrancy_level--;
         return;
     }
@@ -2294,6 +2292,17 @@ static const struct SCSIBusInfo lsi_scsi_info = {
     .cancel = lsi_request_cancelled
 };
 
+static void scripts_timer_cb(void *opaque)
+{
+    LSIState *s = opaque;
+
+    trace_lsi_scripts_timer_triggered();
+    if (s->waiting == LSI_WAIT_SCRIPTS) {
+        s->waiting = 0;
+        lsi_execute_script(s);
+    }
+}
+
 static void lsi_scsi_realize(PCIDevice *dev, Error **errp)
 {
     LSIState *s = LSI53C895A(dev);
@@ -2313,6 +2322,7 @@ static void lsi_scsi_realize(PCIDevice *dev, Error **errp)
                           "lsi-ram", 0x2000);
     memory_region_init_io(&s->io_io, OBJECT(s), &lsi_io_ops, s,
                           "lsi-io", 256);
+    s->scripts_timer = timer_new_us(QEMU_CLOCK_VIRTUAL, scripts_timer_cb, s);
 
     /*
      * Since we use the address-space API to interact with ram_io, disable the
diff --git a/hw/scsi/trace-events b/hw/scsi/trace-events
index d72f741ed8..4456a08ab0 100644
--- a/hw/scsi/trace-events
+++ b/hw/scsi/trace-events
@@ -302,6 +302,8 @@ lsi_execute_script_stop(void) "SCRIPTS execution stopped"
 lsi_awoken(void) "Woken by SIGP"
 lsi_reg_read(const char *name, int offset, uint8_t ret) "Read reg %s 0x%x = 0x%02x"
 lsi_reg_write(const char *name, int offset, uint8_t val) "Write reg %s 0x%x = 0x%02x"
+lsi_scripts_timer_triggered(void) "SCRIPTS timer triggered"
+lsi_scripts_timer_arm(void) "SCRIPTS timer armed"
 
 # virtio-scsi.c
 virtio_scsi_cmd_req(int lun, uint32_t tag, uint8_t cmd) "virtio_scsi_cmd_req lun=%u tag=0x%x cmd=0x%x"


  reply	other threads:[~2024-02-29 16:55 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-28 21:11 [PATCH] hw/scsi/lsi53c895a: add hack to prevent scsi timeouts in HP-UX 10.20 Sven Schnelle
2024-02-29  9:36 ` Peter Maydell
2024-02-29 16:54   ` Sven Schnelle [this message]
2024-02-29 17:26     ` Peter Maydell
2024-02-29 18:01       ` Sven Schnelle

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=87r0gvxlce.fsf@t14.stackframe.org \
    --to=svens@stackframe.org \
    --cc=deller@gmx.de \
    --cc=fam@euphon.net \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    /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).