All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Marek Marczykowski-Górecki" <marmarek@invisiblethingslab.com>
To: xen-devel@lists.xenproject.org
Cc: "Marek Marczykowski-Górecki" <marmarek@invisiblethingslab.com>,
	"Andrew Cooper" <andrew.cooper3@citrix.com>,
	"George Dunlap" <george.dunlap@citrix.com>,
	"Jan Beulich" <jbeulich@suse.com>,
	"Julien Grall" <julien@xen.org>,
	"Stefano Stabellini" <sstabellini@kernel.org>,
	"Wei Liu" <wl@xen.org>
Subject: [RFC PATCH 03/12] xue: reset XHCI ports when initializing dbc
Date: Mon,  6 Jun 2022 05:40:15 +0200	[thread overview]
Message-ID: <27748e5f94769a66900697e521b35b61b1da01d8.1654486751.git-series.marmarek@invisiblethingslab.com> (raw)
In-Reply-To: <cover.07846d9c1900bd8c905a05e7afb214b4cf4ab881.1654486751.git-series.marmarek@invisiblethingslab.com>

Reset ports, to force host system to re-enumerate devices. Otheriwse it
will require the cable to be re-plugged, or will wait in the
"configuring" state indefinitely.

Trick and code copied from Linux:
drivers/usb/early/xhci-dbc.c:xdbc_start()->xdbc_reset_debug_port()

Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com>
---
 xen/include/xue.h | 69 ++++++++++++++++++++++++++++++++++++++++++++++++-
 1 file changed, 69 insertions(+)

diff --git a/xen/include/xue.h b/xen/include/xue.h
index 7515244f6af3..6048dcdd5509 100644
--- a/xen/include/xue.h
+++ b/xen/include/xue.h
@@ -59,6 +59,10 @@
     ((1UL << XUE_PSC_CSC) | (1UL << XUE_PSC_PRC) | (1UL << XUE_PSC_PLC) |      \
      (1UL << XUE_PSC_CEC))
 
+#define     XUE_XHC_EXT_PORT_MAJOR(x)  (((x) >> 24) & 0xff)
+#define PORT_RESET  (1 << 4)
+#define PORT_CONNECT  (1 << 0)
+
 static inline int known_xhc(uint32_t dev_ven)
 {
     switch (dev_ven) {
@@ -1420,6 +1424,67 @@ static inline void xue_init_strings(struct xue *xue, uint32_t *info)
     info[8] = (4 << 24) | (30 << 16) | (8 << 8) | 6;
 }
 
+static inline void xue_do_reset_debug_port(struct xue *xue, u32 id, u32 count)
+{
+    uint32_t *ops_reg;
+    uint32_t *portsc;
+    u32 val, cap_length;
+    int i;
+
+    cap_length = (*(uint32_t*)xue->xhc_mmio) & 0xff;
+    ops_reg = xue->xhc_mmio + cap_length;
+
+    id--;
+    for (i = id; i < (id + count); i++) {
+        portsc = ops_reg + 0x100 + i * 0x4;
+        val = *portsc;
+        if (!(val & PORT_CONNECT))
+            *portsc = val | PORT_RESET;
+    }
+}
+
+
+static inline void xue_reset_debug_port(struct xue *xue)
+{
+    u32 val, port_offset, port_count;
+    uint32_t *xcap;
+    uint32_t next;
+    uint32_t id;
+    uint8_t *mmio = (uint8_t *)xue->xhc_mmio;
+    uint32_t *hccp1 = (uint32_t *)(mmio + 0x10);
+    const uint32_t PROTOCOL_ID = 0x2;
+
+    /**
+     * Paranoid check against a zero value. The spec mandates that
+     * at least one "supported protocol" capability must be implemented,
+     * so this should always be false.
+     */
+    if ((*hccp1 & 0xFFFF0000) == 0) {
+        return;
+    }
+
+    xcap = (uint32_t *)(mmio + (((*hccp1 & 0xFFFF0000) >> 16) << 2));
+    next = (*xcap & 0xFF00) >> 8;
+    id = *xcap & 0xFF;
+
+    /* Look for "supported protocol" capability, major revision 3 */
+    for (;next; xcap += next, id = *xcap & 0xFF, next = (*xcap & 0xFF00) >> 8) {
+        if (id != PROTOCOL_ID && next)
+            continue;
+
+        if (XUE_XHC_EXT_PORT_MAJOR(*xcap) != 0x3)
+            continue;
+
+        /* extract ports offset and count from the capability structure */
+        val = *(xcap + 2);
+        port_offset = val & 0xff;
+        port_count = (val >> 8) & 0xff;
+
+        /* and reset them all */
+        xue_do_reset_debug_port(xue, port_offset, port_count);
+    }
+}
+
 static inline void xue_dump(struct xue *xue)
 {
     struct xue_ops *op = xue->ops;
@@ -1459,6 +1524,10 @@ static inline void xue_enable_dbc(struct xue *xue)
         ops->pause(sys);
     }
 
+    /* reset ports on initial open, to force re-enumerating by the host */
+    if (!xue->open)
+        xue_reset_debug_port(xue);
+
     ops->sfence(sys);
     reg->portsc |= (1UL << XUE_PSC_PED);
     ops->sfence(sys);
-- 
git-series 0.9.1


  parent reply	other threads:[~2022-06-06  3:41 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-06  3:40 [RFC PATCH 00/12] Add Xue - console over USB 3 Debug Capability Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 01/12] drivers/char: Add support for Xue USB3 debugger Marek Marczykowski-Górecki
2022-06-06 13:32   ` Tamas K Lengyel
2022-06-06 14:03     ` Marek Marczykowski-Górecki
2022-06-06 14:10       ` Tamas K Lengyel
2022-06-06 16:57         ` Tamas K Lengyel
2022-06-06 17:04           ` Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 02/12] xue: annotate functions with cf_check Marek Marczykowski-Górecki
2022-06-06  3:40 ` Marek Marczykowski-Górecki [this message]
2022-06-06  3:40 ` [RFC PATCH 04/12] xue: add support for selecting specific xhci Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 05/12] ehci-dbgp: fix selecting n-th ehci controller Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 06/12] console: support multiple serial console simultaneously Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 07/12] IOMMU: add common API for device reserved memory Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 08/12] IOMMU/VT-d: wire common device reserved memory API Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 09/12] IOMMU/AMD: " Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 10/12] xue: mark DMA buffers as reserved for the device Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 11/12] xue: prevent dom0 (or other domain) from using " Marek Marczykowski-Górecki
2022-06-06  3:40 ` [RFC PATCH 12/12] xue: allow driving the reset of XHCI by a domain while Xen uses DbC Marek Marczykowski-Górecki
2022-06-06 13:18 ` [RFC PATCH 00/12] Add Xue - console over USB 3 Debug Capability Lengyel, Tamas
2022-06-06 13:37   ` Marczykowski, Marek
2022-06-06 13:18 ` Andrew Cooper
2022-06-06 13:56   ` Marek Marczykowski-Górecki

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=27748e5f94769a66900697e521b35b61b1da01d8.1654486751.git-series.marmarek@invisiblethingslab.com \
    --to=marmarek@invisiblethingslab.com \
    --cc=andrew.cooper3@citrix.com \
    --cc=george.dunlap@citrix.com \
    --cc=jbeulich@suse.com \
    --cc=julien@xen.org \
    --cc=sstabellini@kernel.org \
    --cc=wl@xen.org \
    --cc=xen-devel@lists.xenproject.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 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.