| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <1551752548.2210.94.camel@mhfsdcap03>
Date: Tue, 5 Mar 2019 10:22:28 +0800
From: Chunfeng Yun <chunfeng.yun@...iatek.com>
To: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
CC: Felipe Balbi <balbi@...nel.org>,
Mathias Nyman <mathias.nyman@...el.com>,
Joel Stanley <joel@....id.au>,
Andrew Jeffery <andrew@...id.au>,
Matthias Brugger <matthias.bgg@...il.com>,
"Yoshihiro Shimoda" <yoshihiro.shimoda.uh@...esas.com>,
Arnd Bergmann <arnd@...db.de>,
Martin Blumenstingl <martin.blumenstingl@...glemail.com>,
Alan Stern <stern@...land.harvard.edu>,
Roger Quadros <rogerq@...com>,
"Sebastian Andrzej Siewior" <bigeasy@...utronix.de>,
"Gustavo A. R. Silva" <gustavo@...eddedor.com>,
Johan Hovold <johan@...nel.org>,
"Benjamin Herrenschmidt" <benh@...nel.crashing.org>,
Colin Ian King <colin.king@...onical.com>,
Jaejoong Kim <climbbb.kim@...il.com>,
<linux-usb@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
<linux-arm-kernel@...ts.infradead.org>,
<linux-aspeed@...ts.ozlabs.org>,
<linux-mediatek@...ts.infradead.org>
Subject: Re: [PATCH] usb: introduce usb_ep_type_string() function
Hi,
On Fri, 2019-03-01 at 17:21 +0100, Greg Kroah-Hartman wrote:
> On Fri, Mar 01, 2019 at 02:58:23PM +0800, Chunfeng Yun wrote:
> > In some places, the code prints a human-readable USB endpoint
> > transfer type (e.g. "bulk"). This involves a switch statement
> > sometimes wrapped around in ({ ... }) block leading to code
> > repetition.
> > To make this scenario easier, here introduces usb_ep_type_string()
> > function, which returns a human-readable name of provided
> > endpoint type.
> > It also changes a few places switch was used to use this
> > new function.
> >
> > Signed-off-by: Chunfeng Yun <chunfeng.yun@...iatek.com>
> > ---
> > drivers/usb/common/common.c | 16 ++++++++++++++++
> > drivers/usb/core/endpoint.c | 18 ++----------------
> > drivers/usb/core/hcd.c | 17 ++---------------
> > drivers/usb/dwc3/debugfs.c | 23 ++++-------------------
> > drivers/usb/gadget/udc/aspeed-vhub/epn.c | 6 +-----
> > drivers/usb/gadget/udc/dummy_hcd.c | 16 +---------------
> > drivers/usb/host/xhci-trace.h | 19 ++-----------------
> > include/linux/usb/ch9.h | 8 ++++++++
> > 8 files changed, 36 insertions(+), 87 deletions(-)
> >
> > diff --git a/drivers/usb/common/common.c b/drivers/usb/common/common.c
> > index 48277bbc15e4..2174dd9ec176 100644
> > --- a/drivers/usb/common/common.c
> > +++ b/drivers/usb/common/common.c
> > @@ -16,6 +16,22 @@
> > #include <linux/usb/otg.h>
> > #include <linux/of_platform.h>
> >
> > +static const char *const ep_type_names[] = {
> > + [USB_ENDPOINT_XFER_CONTROL] = "ctrl",
> > + [USB_ENDPOINT_XFER_ISOC] = "isoc",
> > + [USB_ENDPOINT_XFER_BULK] = "bulk",
> > + [USB_ENDPOINT_XFER_INT] = "intr",
> > +};
> > +
> > +const char *usb_ep_type_string(int ep_type)
> > +{
> > + if (ep_type < 0 || ep_type >= ARRAY_SIZE(ep_type_names))
> > + return "unknown";
> > +
> > + return ep_type_names[ep_type];
> > +}
> > +EXPORT_SYMBOL_GPL(usb_ep_type_string);
> > +
> > const char *usb_otg_state_string(enum usb_otg_state state)
> > {
> > static const char *const names[] = {
> > diff --git a/drivers/usb/core/endpoint.c b/drivers/usb/core/endpoint.c
> > index 1c2c04079676..afa43f9a47b2 100644
> > --- a/drivers/usb/core/endpoint.c
> > +++ b/drivers/usb/core/endpoint.c
> > @@ -60,23 +60,9 @@ static ssize_t type_show(struct device *dev, struct device_attribute *attr,
> > char *buf)
> > {
> > struct ep_device *ep = to_ep_device(dev);
> > - char *type = "unknown";
> > + int ep_type = usb_endpoint_type(ep->desc);
> >
> > - switch (usb_endpoint_type(ep->desc)) {
> > - case USB_ENDPOINT_XFER_CONTROL:
> > - type = "Control";
> > - break;
> > - case USB_ENDPOINT_XFER_ISOC:
> > - type = "Isoc";
> > - break;
> > - case USB_ENDPOINT_XFER_BULK:
> > - type = "Bulk";
> > - break;
> > - case USB_ENDPOINT_XFER_INT:
> > - type = "Interrupt";
> > - break;
> > - }
> > - return sprintf(buf, "%s\n", type);
> > + return sprintf(buf, "%s\n", usb_ep_type_string(ep_type));
>
> You just changed a user/kernel API here, the strings are now different
> from what they used to be :(
How about using these type string to keep them unchanged?
I need help from Felipe and Mathias to check whether the changes in dwc3
and xhci have side effect or not, and waiting for their comments
Thanks
>
> That's not ok, odds are you will break tools if you do this.
>
> > }
> > static DEVICE_ATTR_RO(type);
> >
> > diff --git a/drivers/usb/core/hcd.c b/drivers/usb/core/hcd.c
> > index 015b126ce455..193ee92b2fdb 100644
> > --- a/drivers/usb/core/hcd.c
> > +++ b/drivers/usb/core/hcd.c
> > @@ -1875,23 +1875,10 @@ void usb_hcd_flush_endpoint(struct usb_device *udev,
> > /* kick hcd */
> > unlink1(hcd, urb, -ESHUTDOWN);
> > dev_dbg (hcd->self.controller,
> > - "shutdown urb %pK ep%d%s%s\n",
> > + "shutdown urb %pK ep%d%s-%s\n",
> > urb, usb_endpoint_num(&ep->desc),
> > is_in ? "in" : "out",
> > - ({ char *s;
> > -
> > - switch (usb_endpoint_type(&ep->desc)) {
> > - case USB_ENDPOINT_XFER_CONTROL:
> > - s = ""; break;
> > - case USB_ENDPOINT_XFER_BULK:
> > - s = "-bulk"; break;
> > - case USB_ENDPOINT_XFER_INT:
> > - s = "-intr"; break;
> > - default:
> > - s = "-iso"; break;
> > - };
> > - s;
> > - }));
> > + usb_ep_type_string(usb_endpoint_type(&ep->desc)));
>
> You also changed the message here for control endpoints, but that's not
> a big deal, it's ok.
>
> > usb_put_urb (urb);
> >
> > /* list contents may have changed */
> > diff --git a/drivers/usb/dwc3/debugfs.c b/drivers/usb/dwc3/debugfs.c
> > index 1c792710348f..d9e2a63835fe 100644
> > --- a/drivers/usb/dwc3/debugfs.c
> > +++ b/drivers/usb/dwc3/debugfs.c
> > @@ -750,28 +750,13 @@ static int dwc3_transfer_type_show(struct seq_file *s, void *unused)
> > unsigned long flags;
> >
> > spin_lock_irqsave(&dwc->lock, flags);
> > - if (!(dep->flags & DWC3_EP_ENABLED) ||
> > - !dep->endpoint.desc) {
> > - seq_printf(s, "--\n");
> > + if (!(dep->flags & DWC3_EP_ENABLED) || !dep->endpoint.desc) {
> > + seq_puts(s, "unknown\n");
> > goto out;
> > }
> >
> > - switch (usb_endpoint_type(dep->endpoint.desc)) {
> > - case USB_ENDPOINT_XFER_CONTROL:
> > - seq_printf(s, "control\n");
> > - break;
> > - case USB_ENDPOINT_XFER_ISOC:
> > - seq_printf(s, "isochronous\n");
> > - break;
> > - case USB_ENDPOINT_XFER_BULK:
> > - seq_printf(s, "bulk\n");
> > - break;
> > - case USB_ENDPOINT_XFER_INT:
> > - seq_printf(s, "interrupt\n");
> > - break;
> > - default:
> > - seq_printf(s, "--\n");
> > - }
> > + seq_printf(s, "%s\n",
> > + usb_ep_type_string(usb_endpoint_type(dep->endpoint.desc)));
>
> It is debugfs, so no real API guarantees at all, but you did change the
> output format, and that's not always good.
got it
>
> >
> > out:
> > spin_unlock_irqrestore(&dwc->lock, flags);
> > diff --git a/drivers/usb/gadget/udc/aspeed-vhub/epn.c b/drivers/usb/gadget/udc/aspeed-vhub/epn.c
> > index 4a28e3fbeb0b..165b0a43670e 100644
> > --- a/drivers/usb/gadget/udc/aspeed-vhub/epn.c
> > +++ b/drivers/usb/gadget/udc/aspeed-vhub/epn.c
> > @@ -593,10 +593,6 @@ static int ast_vhub_epn_disable(struct usb_ep* u_ep)
> > static int ast_vhub_epn_enable(struct usb_ep* u_ep,
> > const struct usb_endpoint_descriptor *desc)
> > {
> > - static const char *ep_type_string[] __maybe_unused = { "ctrl",
> > - "isoc",
> > - "bulk",
> > - "intr" };
> > struct ast_vhub_ep *ep = to_ast_ep(u_ep);
> > struct ast_vhub_dev *dev;
> > struct ast_vhub *vhub;
> > @@ -646,7 +642,7 @@ static int ast_vhub_epn_enable(struct usb_ep* u_ep,
> > ep->epn.wedged = false;
> >
> > EPDBG(ep, "Enabling [%s] %s num %d maxpacket=%d\n",
> > - ep->epn.is_in ? "in" : "out", ep_type_string[type],
> > + ep->epn.is_in ? "in" : "out", usb_ep_type_string(type),
> > usb_endpoint_num(desc), maxpacket);
> >
> > /* Can we use DMA descriptor mode ? */
> > diff --git a/drivers/usb/gadget/udc/dummy_hcd.c b/drivers/usb/gadget/udc/dummy_hcd.c
> > index baf72f95f0f1..40c6a484e232 100644
> > --- a/drivers/usb/gadget/udc/dummy_hcd.c
> > +++ b/drivers/usb/gadget/udc/dummy_hcd.c
> > @@ -617,21 +617,7 @@ static int dummy_enable(struct usb_ep *_ep,
> > _ep->name,
> > desc->bEndpointAddress & 0x0f,
> > (desc->bEndpointAddress & USB_DIR_IN) ? "in" : "out",
> > - ({ char *val;
> > - switch (usb_endpoint_type(desc)) {
> > - case USB_ENDPOINT_XFER_BULK:
> > - val = "bulk";
> > - break;
> > - case USB_ENDPOINT_XFER_ISOC:
> > - val = "iso";
> > - break;
> > - case USB_ENDPOINT_XFER_INT:
> > - val = "intr";
> > - break;
> > - default:
> > - val = "ctrl";
> > - break;
> > - } val; }),
> > + usb_ep_type_string(usb_endpoint_type(desc)),
> > max, ep->stream_en ? "enabled" : "disabled");
> >
> > /* at this point real hardware should be NAKing transfers
> > diff --git a/drivers/usb/host/xhci-trace.h b/drivers/usb/host/xhci-trace.h
> > index 88b427434bd8..bac73d688f11 100644
> > --- a/drivers/usb/host/xhci-trace.h
> > +++ b/drivers/usb/host/xhci-trace.h
> > @@ -289,23 +289,8 @@ DECLARE_EVENT_CLASS(xhci_log_urb,
> > ),
> > TP_printk("ep%d%s-%s: urb %p pipe %u slot %d length %d/%d sgs %d/%d stream %d flags %08x",
> > __entry->epnum, __entry->dir_in ? "in" : "out",
> > - ({ char *s;
> > - switch (__entry->type) {
> > - case USB_ENDPOINT_XFER_INT:
> > - s = "intr";
> > - break;
> > - case USB_ENDPOINT_XFER_CONTROL:
> > - s = "control";
> > - break;
> > - case USB_ENDPOINT_XFER_BULK:
> > - s = "bulk";
> > - break;
> > - case USB_ENDPOINT_XFER_ISOC:
> > - s = "isoc";
> > - break;
> > - default:
> > - s = "UNKNOWN";
> > - } s; }), __entry->urb, __entry->pipe, __entry->slot_id,
> > + usb_ep_type_string(__entry->type),
> > + __entry->urb, __entry->pipe, __entry->slot_id,
> > __entry->actual, __entry->length, __entry->num_mapped_sgs,
> > __entry->num_sgs, __entry->stream, __entry->flags
> > )
>
> Another api change with "control" :(
>
> I like the idea, but you can not break user/kernel apis like this
> without knowing the affects that could happen.
>
> thanks,
>
> greg k-h
Powered by blists - more mailing lists