[Qemu-devel] [PATCH v2 0/1] ide-test: tidy up after pio race fix

[Qemu-devel] [PATCH v2 0/1] ide-test: tidy up after pio race fix

[John Snow]

Post-rc1 cleanup of the ATAPI tests that were problematic during the
rc0 testing window. This set does two things:

(1) Fix the timeouts to be more deterministic, and
(2) Finish tidying up the PIO loop.

v2:
 - Clarify the commit message on patch 1, and remove the second
   needless stack variable.

John Snow (1):
  ide-test: fix timeouts

 tests/ide-test.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)

-- 
2.4.3

[Qemu-devel] [PATCH v2 1/1] ide-test: fix timeouts

[John Snow]

Use explicit timeouts instead of trying to approximate it by counting
the cumulative duration of nsleep calls.

In practice, the timeout if inb() dwarfed the nsleep delays, and as a
result the real timeout value became a lot larger than 5 seconds.

So: change the semantics from "Not sooner than 5 seconds" to "no more
than 5 seconds" to ensure we don't hang the tester for very long.

Signed-off-by: John Snow <jsnow@redhat.com>
---
 tests/ide-test.c | 16 ++++++++++++----
 1 file changed, 12 insertions(+), 4 deletions(-)

diff --git a/tests/ide-test.c b/tests/ide-test.c
index fc1ce52..c14910b 100644
--- a/tests/ide-test.c
+++ b/tests/ide-test.c
@@ -642,15 +642,19 @@ static void nsleep(int64_t nsecs)
 
 static uint8_t ide_wait_clear(uint8_t flag)
 {
-    int i;
     uint8_t data;
+    time_t st;
 
     /* Wait with a 5 second timeout */
-    for (i = 0; i <= 12500000; i++) {
+    time(&st);
+    while (true) {
         data = inb(IDE_BASE + reg_status);
         if (!(data & flag)) {
             return data;
         }
+        if (difftime(time(NULL), st) > 5.0) {
+            break;
+        }
         nsleep(400);
     }
     g_assert_not_reached();
@@ -658,14 +662,18 @@ static uint8_t ide_wait_clear(uint8_t flag)
 
 static void ide_wait_intr(int irq)
 {
-    int i;
+    time_t st;
     bool intr;
 
-    for (i = 0; i <= 12500000; i++) {
+    time(&st);
+    while (true) {
         intr = get_irq(irq);
         if (intr) {
             return;
         }
+        if (difftime(time(NULL), st) > 5.0) {
+            break;
+        }
         nsleep(400);
     }
 
-- 
2.4.3

Re: [Qemu-devel] [PATCH v2 1/1] ide-test: fix timeouts

[Kevin Wolf]

Am 24.11.2015 um 20:36 hat John Snow geschrieben:
> Use explicit timeouts instead of trying to approximate it by counting
> the cumulative duration of nsleep calls.
> 
> In practice, the timeout if inb() dwarfed the nsleep delays, and as a
> result the real timeout value became a lot larger than 5 seconds.
> 
> So: change the semantics from "Not sooner than 5 seconds" to "no more
> than 5 seconds" to ensure we don't hang the tester for very long.
> 
> Signed-off-by: John Snow <jsnow@redhat.com>

Reviewed-by: Kevin Wolf <kwolf@redhat.com>

Re: [Qemu-devel] [PATCH v2 0/1] ide-test: tidy up after pio race fix

[John Snow]

On 11/24/2015 02:36 PM, John Snow wrote:
> Post-rc1 cleanup of the ATAPI tests that were problematic during the
> rc0 testing window. This set does two things:
> 
> (1) Fix the timeouts to be more deterministic, and
> (2) Finish tidying up the PIO loop.
> 
> v2:
>  - Clarify the commit message on patch 1, and remove the second
>    needless stack variable.
> 
> John Snow (1):
>   ide-test: fix timeouts
> 
>  tests/ide-test.c | 16 ++++++++++++----
>  1 file changed, 12 insertions(+), 4 deletions(-)
> 

Script goofed up and didn't send the second one, but since Kevin already
reviewed it, I'll just stage it.

--js