From: Nicholas Piggin <npiggin@gmail.com>
To: qemu-devel@nongnu.org
Cc: Nicholas Piggin <npiggin@gmail.com>,
"Michael S. Tsirkin" <mst@redhat.com>,
Marcel Apfelbaum <marcel.apfelbaum@gmail.com>,
Fabiano Rosas <farosas@suse.de>,
Laurent Vivier <lvivier@redhat.com>,
Paolo Bonzini <pbonzini@redhat.com>,
Dmitry Fleytman <dmitry.fleytman@gmail.com>,
Akihiko Odaki <akihiko.odaki@daynix.com>,
Sriram Yagnaraman <sriram.yagnaraman@ericsson.com>
Subject: [PATCH 4/8] tests/qtest/e1000e|igb: Fix e1000e and igb tests to re-trigger interrupts
Date: Thu, 12 Dec 2024 18:34:57 +1000 [thread overview]
Message-ID: <20241212083502.1439033-5-npiggin@gmail.com> (raw)
In-Reply-To: <20241212083502.1439033-1-npiggin@gmail.com>
The e1000e and igb tests don't clear the msix pending bit after waiting
for it sit is masked so the irq doesn't get delivered. Failing to clear
the pending interrupt means all subsequent waits for interrupt after the
first do not actually wait for an interrupt genreated by the device.
Explicitly clearing the msix pending bit results in the
multiple-transfers test hanging waiting for the second interrupt. This
happens because the e1000e and igb tests do not clear (or set
auto-clear) on queue interrupts, so the cause remains ste in ICR/EICR,
which inhibits triggering of a new interrupt.
Fix both these problems. Clear the msix pending bit explicitly after
waiting for it; and clear the ICR/EICR cause bits after seeing and
interrupt (also verify we saw the correct cause bit).
Cc: Michael S. Tsirkin <mst@redhat.com>
Cc: Marcel Apfelbaum <marcel.apfelbaum@gmail.com>
Cc: Dmitry Fleytman <dmitry.fleytman@gmail.com>
Cc: Akihiko Odaki <akihiko.odaki@daynix.com>
Cc: Sriram Yagnaraman <sriram.yagnaraman@ericsson.com>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
tests/qtest/e1000e-test.c | 8 ++++++--
tests/qtest/igb-test.c | 8 ++++++--
tests/qtest/libqos/e1000e.c | 2 +-
3 files changed, 13 insertions(+), 5 deletions(-)
diff --git a/tests/qtest/e1000e-test.c b/tests/qtest/e1000e-test.c
index de9738fdb74..a69759da70e 100644
--- a/tests/qtest/e1000e-test.c
+++ b/tests/qtest/e1000e-test.c
@@ -64,8 +64,10 @@ static void e1000e_send_verify(QE1000E *d, int *test_sockets, QGuestAllocator *a
/* Put descriptor to the ring */
e1000e_tx_ring_push(d, &descr);
- /* Wait for TX WB interrupt */
+ /* Wait for TX WB interrupt (this clears the MSIX PBA) */
e1000e_wait_isr(d, E1000E_TX0_MSG_ID);
+ /* Read ICR which clears it ready for next interrupt, assert TXQ0 cause */
+ g_assert(e1000e_macreg_read(d, E1000_ICR) & E1000_ICR_TXQ0);
/* Check DD bit */
g_assert_cmphex(le32_to_cpu(descr.upper.data) & E1000_TXD_STAT_DD, ==,
@@ -115,8 +117,10 @@ static void e1000e_receive_verify(QE1000E *d, int *test_sockets, QGuestAllocator
/* Put descriptor to the ring */
e1000e_rx_ring_push(d, &descr);
- /* Wait for TX WB interrupt */
+ /* Wait for TX WB interrupt (this clears the MSIX PBA) */
e1000e_wait_isr(d, E1000E_RX0_MSG_ID);
+ /* Read ICR which clears it ready for next interrupt, assert RXQ0 cause */
+ g_assert(e1000e_macreg_read(d, E1000_ICR) & E1000_ICR_RXQ0);
/* Check DD bit */
g_assert_cmphex(le32_to_cpu(descr.wb.upper.status_error) &
diff --git a/tests/qtest/igb-test.c b/tests/qtest/igb-test.c
index 3d397ea6973..2f22c4fb208 100644
--- a/tests/qtest/igb-test.c
+++ b/tests/qtest/igb-test.c
@@ -67,8 +67,10 @@ static void igb_send_verify(QE1000E *d, int *test_sockets, QGuestAllocator *allo
/* Put descriptor to the ring */
e1000e_tx_ring_push(d, &descr);
- /* Wait for TX WB interrupt */
+ /* Wait for TX WB interrupt (this clears the MSIX PBA) */
e1000e_wait_isr(d, E1000E_TX0_MSG_ID);
+ /* Read EICR which clears it ready for next interrupt, assert TXQ0 cause */
+ g_assert(e1000e_macreg_read(d, E1000_EICR) & (1 << E1000E_TX0_MSG_ID));
/* Check DD bit */
g_assert_cmphex(le32_to_cpu(descr.wb.status) & E1000_TXD_STAT_DD, ==,
@@ -118,8 +120,10 @@ static void igb_receive_verify(QE1000E *d, int *test_sockets, QGuestAllocator *a
/* Put descriptor to the ring */
e1000e_rx_ring_push(d, &descr);
- /* Wait for TX WB interrupt */
+ /* Wait for TX WB interrupt (this clears the MSIX PBA) */
e1000e_wait_isr(d, E1000E_RX0_MSG_ID);
+ /* Read EICR which clears it ready for next interrupt, assert RXQ0 cause */
+ g_assert(e1000e_macreg_read(d, E1000_EICR) & (1 << E1000E_RX0_MSG_ID));
/* Check DD bit */
g_assert_cmphex(le32_to_cpu(descr.wb.upper.status_error) &
diff --git a/tests/qtest/libqos/e1000e.c b/tests/qtest/libqos/e1000e.c
index 925654c7fd4..8ef6a04f43e 100644
--- a/tests/qtest/libqos/e1000e.c
+++ b/tests/qtest/libqos/e1000e.c
@@ -83,7 +83,7 @@ void e1000e_wait_isr(QE1000E *d, uint16_t msg_id)
guint64 end_time = g_get_monotonic_time() + 5 * G_TIME_SPAN_SECOND;
do {
- if (qpci_msix_pending(&d_pci->pci_dev, msg_id)) {
+ if (qpci_msix_test_clear_pending(&d_pci->pci_dev, msg_id)) {
return;
}
qtest_clock_step(d_pci->pci_dev.bus->qts, 10000);
--
2.45.2
next prev parent reply other threads:[~2024-12-12 8:36 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-12-12 8:34 [PATCH 0/8] Add XHCI TR NOOP support, plus PCI, MSIX changes Nicholas Piggin
2024-12-12 8:34 ` [PATCH 1/8] qtest/pci: Enforce balanced iomap/unmap Nicholas Piggin
2024-12-12 8:34 ` [PATCH 2/8] qtest/libqos/pci: Fix qpci_msix_enable sharing bar0 Nicholas Piggin
2024-12-12 8:34 ` [PATCH 3/8] pci/msix: Implement PBA writes Nicholas Piggin
2024-12-13 5:14 ` Akihiko Odaki
2024-12-18 1:44 ` Nicholas Piggin
2024-12-12 8:34 ` Nicholas Piggin [this message]
2024-12-12 8:34 ` [PATCH 5/8] hw/usb/xhci: Move HCD constants to a header and add register constants Nicholas Piggin
2024-12-18 15:08 ` Phil Dennis-Jordan
2024-12-19 1:50 ` Nicholas Piggin
2024-12-20 14:11 ` Phil Dennis-Jordan
2024-12-12 8:34 ` [PATCH 6/8] qtest/xhci: Add controller and device setup and ring tests Nicholas Piggin
2024-12-12 8:35 ` [PATCH 7/8] hw/usb/xhci: Support TR NOOP commands Nicholas Piggin
2024-12-12 8:35 ` [PATCH 8/8] qtest/xhci: add a test for " Nicholas Piggin
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=20241212083502.1439033-5-npiggin@gmail.com \
--to=npiggin@gmail.com \
--cc=akihiko.odaki@daynix.com \
--cc=dmitry.fleytman@gmail.com \
--cc=farosas@suse.de \
--cc=lvivier@redhat.com \
--cc=marcel.apfelbaum@gmail.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=sriram.yagnaraman@ericsson.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.