* [PATCH RESEND v7 0/2] usb: gadget: f_fs: userspace API fixes and improvements @ 2014-09-08 7:56 Robert Baldyga 2014-09-08 7:56 ` [PATCH RESEND v7 1/2] usb: gadget: f_fs: add ioctl returning ep descriptor Robert Baldyga 2014-09-08 7:57 ` [PATCH RESEND v7 2/2] usb: gadget: f_fs: virtual endpoint address mapping Robert Baldyga 0 siblings, 2 replies; 6+ messages in thread From: Robert Baldyga @ 2014-09-08 7:56 UTC (permalink / raw) To: balbi; +Cc: gregkh, linux-usb, linux-kernel, Robert Baldyga Hi Felipe, I have rebased this patchset at your testing/next/ branch. Best regards Robert Baldyga Robert Baldyga (2): usb: gadget: f_fs: add ioctl returning ep descriptor usb: gadget: f_fs: virtual endpoint address mapping drivers/usb/gadget/function/f_fs.c | 46 +++++++++++++++++++++++++++++++++++-- drivers/usb/gadget/function/u_fs.h | 2 ++ include/uapi/linux/usb/functionfs.h | 7 ++++++ 3 files changed, 53 insertions(+), 2 deletions(-) -- 1.9.1 ^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH RESEND v7 1/2] usb: gadget: f_fs: add ioctl returning ep descriptor 2014-09-08 7:56 [PATCH RESEND v7 0/2] usb: gadget: f_fs: userspace API fixes and improvements Robert Baldyga @ 2014-09-08 7:56 ` Robert Baldyga 2014-09-08 7:57 ` [PATCH RESEND v7 2/2] usb: gadget: f_fs: virtual endpoint address mapping Robert Baldyga 1 sibling, 0 replies; 6+ messages in thread From: Robert Baldyga @ 2014-09-08 7:56 UTC (permalink / raw) To: balbi; +Cc: gregkh, linux-usb, linux-kernel, Robert Baldyga This patch introduces ioctl named FUNCTIONFS_ENDPOINT_DESC, which returns endpoint descriptor to userspace. It works only if function is active. Signed-off-by: Robert Baldyga <r.baldyga@samsung.com> Acked-by: Michal Nazarewicz <mina86@mina86.com> --- drivers/usb/gadget/function/f_fs.c | 23 +++++++++++++++++++++++ include/uapi/linux/usb/functionfs.h | 6 ++++++ 2 files changed, 29 insertions(+) diff --git a/drivers/usb/gadget/function/f_fs.c b/drivers/usb/gadget/function/f_fs.c index dc30adf..6405e0f 100644 --- a/drivers/usb/gadget/function/f_fs.c +++ b/drivers/usb/gadget/function/f_fs.c @@ -1026,6 +1026,29 @@ static long ffs_epfile_ioctl(struct file *file, unsigned code, case FUNCTIONFS_ENDPOINT_REVMAP: ret = epfile->ep->num; break; + case FUNCTIONFS_ENDPOINT_DESC: + { + int desc_idx; + struct usb_endpoint_descriptor *desc; + + switch (epfile->ffs->gadget->speed) { + case USB_SPEED_SUPER: + desc_idx = 2; + break; + case USB_SPEED_HIGH: + desc_idx = 1; + break; + default: + desc_idx = 0; + } + desc = epfile->ep->descs[desc_idx]; + + spin_unlock_irq(&epfile->ffs->eps_lock); + ret = copy_to_user((void *)value, desc, sizeof(*desc)); + if (ret) + ret = -EFAULT; + return ret; + } default: ret = -ENOTTY; } diff --git a/include/uapi/linux/usb/functionfs.h b/include/uapi/linux/usb/functionfs.h index 6d2a16b..7c7a2fe 100644 --- a/include/uapi/linux/usb/functionfs.h +++ b/include/uapi/linux/usb/functionfs.h @@ -275,6 +275,12 @@ struct usb_functionfs_event { */ #define FUNCTIONFS_ENDPOINT_REVMAP _IO('g', 129) +/* + * Returns endpoint descriptor. If function is not active returns -ENODEV. + */ +#define FUNCTIONFS_ENDPOINT_DESC _IOR('g', 130, \ + struct usb_endpoint_descriptor) + #endif /* _UAPI__LINUX_FUNCTIONFS_H__ */ -- 1.9.1 ^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH RESEND v7 2/2] usb: gadget: f_fs: virtual endpoint address mapping 2014-09-08 7:56 [PATCH RESEND v7 0/2] usb: gadget: f_fs: userspace API fixes and improvements Robert Baldyga 2014-09-08 7:56 ` [PATCH RESEND v7 1/2] usb: gadget: f_fs: add ioctl returning ep descriptor Robert Baldyga @ 2014-09-08 7:57 ` Robert Baldyga 2014-09-08 13:47 ` Felipe Balbi 1 sibling, 1 reply; 6+ messages in thread From: Robert Baldyga @ 2014-09-08 7:57 UTC (permalink / raw) To: balbi; +Cc: gregkh, linux-usb, linux-kernel, Robert Baldyga This patch introduces virtual endpoint address mapping. It separates function logic form physical endpoint addresses making it more hardware independent. Following modifications changes user space API, so to enable them user have to switch on the FUNCTIONFS_VIRTUAL_ADDR flag in descriptors. Endpoints are now refered using virtual endpoint addresses chosen by user in endpoint descpriptors. This applies to each context when endpoint address can be used: - when accessing endpoint files in FunctionFS filesystemi (in file name), - in setup requests directed to specific endpoint (in wIndex field), - in descriptors returned by FUNCTIONFS_ENDPOINT_DESC ioctl. In endpoint file names the endpoint address number is formatted as double-digit hexadecimal value ("ep%02x") which has few advantages - it is easy to parse, allows to easly recognize endpoint direction basing on its name (IN endpoint number starts with digit 8, and OUT with 0) which can be useful for debugging purpose, and it makes easier to introduce further features allowing to use each endpoint number in both directions to have more endpoints available for function if hardware supports this (for example we could have ep01 which is endpoint 1 with OUT direction, and ep81 which is endpoint 1 with IN direction). Physical endpoint address can be still obtained using ioctl named FUNCTIONFS_ENDPOINT_REVMAP, but now it's not neccesary to handle USB transactions properly. Signed-off-by: Robert Baldyga <r.baldyga@samsung.com> Acked-by: Michal Nazarewicz <mina86@mina86.com> --- drivers/usb/gadget/function/f_fs.c | 23 +++++++++++++++++++++-- drivers/usb/gadget/function/u_fs.h | 2 ++ include/uapi/linux/usb/functionfs.h | 1 + 3 files changed, 24 insertions(+), 2 deletions(-) diff --git a/drivers/usb/gadget/function/f_fs.c b/drivers/usb/gadget/function/f_fs.c index 6405e0f..071683d 100644 --- a/drivers/usb/gadget/function/f_fs.c +++ b/drivers/usb/gadget/function/f_fs.c @@ -1551,7 +1551,10 @@ static int ffs_epfiles_create(struct ffs_data *ffs) epfile->ffs = ffs; mutex_init(&epfile->mutex); init_waitqueue_head(&epfile->wait); - sprintf(epfiles->name, "ep%u", i); + if (ffs->user_flags & FUNCTIONFS_VIRTUAL_ADDR) + sprintf(epfiles->name, "ep%02x", ffs->eps_addrmap[i]); + else + sprintf(epfiles->name, "ep%u", i); if (!unlikely(ffs_sb_create_file(ffs->sb, epfiles->name, epfile, &ffs_epfile_operations, &epfile->dentry))) { @@ -2090,10 +2093,12 @@ static int __ffs_data_got_descs(struct ffs_data *ffs, break; case FUNCTIONFS_DESCRIPTORS_MAGIC_V2: flags = get_unaligned_le32(data + 8); + ffs->user_flags = flags; if (flags & ~(FUNCTIONFS_HAS_FS_DESC | FUNCTIONFS_HAS_HS_DESC | FUNCTIONFS_HAS_SS_DESC | - FUNCTIONFS_HAS_MS_OS_DESC)) { + FUNCTIONFS_HAS_MS_OS_DESC | + FUNCTIONFS_VIRTUAL_ADDR)) { ret = -ENOSYS; goto error; } @@ -2420,7 +2425,13 @@ static int __ffs_func_bind_do_descs(enum ffs_entity_type type, u8 *valuep, } else { struct usb_request *req; struct usb_ep *ep; + u8 bEndpointAddress; + /* + * We back up bEndpointAddress because autoconfig overwrites + * it with physical endpoint address. + */ + bEndpointAddress = ds->bEndpointAddress; pr_vdebug("autoconfig\n"); ep = usb_ep_autoconfig(func->gadget, ds); if (unlikely(!ep)) @@ -2435,6 +2446,12 @@ static int __ffs_func_bind_do_descs(enum ffs_entity_type type, u8 *valuep, ffs_ep->req = req; func->eps_revmap[ds->bEndpointAddress & USB_ENDPOINT_NUMBER_MASK] = idx + 1; + /* + * If we use virtual address mapping, we restore + * original bEndpointAddress value. + */ + if (func->ffs->user_flags & FUNCTIONFS_VIRTUAL_ADDR) + ds->bEndpointAddress = bEndpointAddress; } ffs_dump_mem(": Rewritten ep desc", ds, ds->bLength); @@ -2879,6 +2896,8 @@ static int ffs_func_setup(struct usb_function *f, ret = ffs_func_revmap_ep(func, le16_to_cpu(creq->wIndex)); if (unlikely(ret < 0)) return ret; + if (func->ffs->user_flags & FUNCTIONFS_VIRTUAL_ADDR) + ret = func->ffs->eps_addrmap[ret]; break; default: diff --git a/drivers/usb/gadget/function/u_fs.h b/drivers/usb/gadget/function/u_fs.h index 63d6e71..bc5a37d 100644 --- a/drivers/usb/gadget/function/u_fs.h +++ b/drivers/usb/gadget/function/u_fs.h @@ -224,6 +224,8 @@ struct ffs_data { void *ms_os_descs_ext_prop_name_avail; void *ms_os_descs_ext_prop_data_avail; + unsigned user_flags; + unsigned short strings_count; unsigned short interfaces_count; unsigned short eps_count; diff --git a/include/uapi/linux/usb/functionfs.h b/include/uapi/linux/usb/functionfs.h index 7c7a2fe..295ba29 100644 --- a/include/uapi/linux/usb/functionfs.h +++ b/include/uapi/linux/usb/functionfs.h @@ -19,6 +19,7 @@ enum functionfs_flags { FUNCTIONFS_HAS_HS_DESC = 2, FUNCTIONFS_HAS_SS_DESC = 4, FUNCTIONFS_HAS_MS_OS_DESC = 8, + FUNCTIONFS_VIRTUAL_ADDR = 16, }; /* Descriptor of an non-audio endpoint */ -- 1.9.1 ^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH RESEND v7 2/2] usb: gadget: f_fs: virtual endpoint address mapping 2014-09-08 7:57 ` [PATCH RESEND v7 2/2] usb: gadget: f_fs: virtual endpoint address mapping Robert Baldyga @ 2014-09-08 13:47 ` Felipe Balbi 2014-09-09 6:02 ` Robert Baldyga 0 siblings, 1 reply; 6+ messages in thread From: Felipe Balbi @ 2014-09-08 13:47 UTC (permalink / raw) To: Robert Baldyga; +Cc: balbi, gregkh, linux-usb, linux-kernel [-- Attachment #1: Type: text/plain, Size: 2578 bytes --] Hi, On Mon, Sep 08, 2014 at 09:57:00AM +0200, Robert Baldyga wrote: > This patch introduces virtual endpoint address mapping. It separates > function logic form physical endpoint addresses making it more hardware > independent. > > Following modifications changes user space API, so to enable them user > have to switch on the FUNCTIONFS_VIRTUAL_ADDR flag in descriptors. > > Endpoints are now refered using virtual endpoint addresses chosen by > user in endpoint descpriptors. This applies to each context when endpoint > address can be used: > - when accessing endpoint files in FunctionFS filesystemi (in file name), > - in setup requests directed to specific endpoint (in wIndex field), > - in descriptors returned by FUNCTIONFS_ENDPOINT_DESC ioctl. > > In endpoint file names the endpoint address number is formatted as > double-digit hexadecimal value ("ep%02x") which has few advantages - > it is easy to parse, allows to easly recognize endpoint direction basing > on its name (IN endpoint number starts with digit 8, and OUT with 0) > which can be useful for debugging purpose, and it makes easier to introduce > further features allowing to use each endpoint number in both directions > to have more endpoints available for function if hardware supports this > (for example we could have ep01 which is endpoint 1 with OUT direction, > and ep81 which is endpoint 1 with IN direction). > > Physical endpoint address can be still obtained using ioctl named > FUNCTIONFS_ENDPOINT_REVMAP, but now it's not neccesary to handle > USB transactions properly. > > Signed-off-by: Robert Baldyga <r.baldyga@samsung.com> > Acked-by: Michal Nazarewicz <mina86@mina86.com> after this patch I get build errors: drivers/usb/gadget/function/f_fs.c: In function ‘ffs_epfiles_create’: drivers/usb/gadget/function/f_fs.c:1555:40: error: ‘struct ffs_data’ has no member named ‘eps_addrmap’ sprintf(epfiles->name, "ep%02x", ffs->eps_addrmap[i]); ^ drivers/usb/gadget/function/f_fs.c: In function ‘ffs_func_setup’: drivers/usb/gadget/function/f_fs.c:2900:19: error: ‘struct ffs_data’ has no member named ‘eps_addrmap’ ret = func->ffs->eps_addrmap[ret]; ^ make[3]: *** [drivers/usb/gadget/function/f_fs.o] Error 1 make[3]: *** Waiting for unfinished jobs.... make[2]: *** [drivers/usb/gadget/function] Error 2 make[1]: *** [drivers/usb/gadget] Error 2 make[1]: *** Waiting for unfinished jobs.... make: *** [drivers/usb/] Error 2 -- balbi [-- Attachment #2: Digital signature --] [-- Type: application/pgp-signature, Size: 819 bytes --] ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH RESEND v7 2/2] usb: gadget: f_fs: virtual endpoint address mapping 2014-09-08 13:47 ` Felipe Balbi @ 2014-09-09 6:02 ` Robert Baldyga 2014-09-09 14:43 ` Felipe Balbi 0 siblings, 1 reply; 6+ messages in thread From: Robert Baldyga @ 2014-09-09 6:02 UTC (permalink / raw) To: balbi; +Cc: gregkh, linux-usb, linux-kernel On 09/08/2014 03:47 PM, Felipe Balbi wrote: > Hi, > > On Mon, Sep 08, 2014 at 09:57:00AM +0200, Robert Baldyga wrote: >> This patch introduces virtual endpoint address mapping. It separates >> function logic form physical endpoint addresses making it more hardware >> independent. >> >> Following modifications changes user space API, so to enable them user >> have to switch on the FUNCTIONFS_VIRTUAL_ADDR flag in descriptors. >> >> Endpoints are now refered using virtual endpoint addresses chosen by >> user in endpoint descpriptors. This applies to each context when endpoint >> address can be used: >> - when accessing endpoint files in FunctionFS filesystemi (in file name), >> - in setup requests directed to specific endpoint (in wIndex field), >> - in descriptors returned by FUNCTIONFS_ENDPOINT_DESC ioctl. >> >> In endpoint file names the endpoint address number is formatted as >> double-digit hexadecimal value ("ep%02x") which has few advantages - >> it is easy to parse, allows to easly recognize endpoint direction basing >> on its name (IN endpoint number starts with digit 8, and OUT with 0) >> which can be useful for debugging purpose, and it makes easier to introduce >> further features allowing to use each endpoint number in both directions >> to have more endpoints available for function if hardware supports this >> (for example we could have ep01 which is endpoint 1 with OUT direction, >> and ep81 which is endpoint 1 with IN direction). >> >> Physical endpoint address can be still obtained using ioctl named >> FUNCTIONFS_ENDPOINT_REVMAP, but now it's not neccesary to handle >> USB transactions properly. >> >> Signed-off-by: Robert Baldyga <r.baldyga@samsung.com> >> Acked-by: Michal Nazarewicz <mina86@mina86.com> > > after this patch I get build errors: > > drivers/usb/gadget/function/f_fs.c: In function ‘ffs_epfiles_create’: > drivers/usb/gadget/function/f_fs.c:1555:40: error: ‘struct ffs_data’ has no member named ‘eps_addrmap’ > sprintf(epfiles->name, "ep%02x", ffs->eps_addrmap[i]); > ^ > drivers/usb/gadget/function/f_fs.c: In function ‘ffs_func_setup’: > drivers/usb/gadget/function/f_fs.c:2900:19: error: ‘struct ffs_data’ has no member named ‘eps_addrmap’ > ret = func->ffs->eps_addrmap[ret]; > ^ > make[3]: *** [drivers/usb/gadget/function/f_fs.o] Error 1 > make[3]: *** Waiting for unfinished jobs.... > make[2]: *** [drivers/usb/gadget/function] Error 2 > make[1]: *** [drivers/usb/gadget] Error 2 > make[1]: *** Waiting for unfinished jobs.... > make: *** [drivers/usb/] Error 2 > Array "eps_addrmap" was introduced in patch "usb: gadget: f_fs: fix the redundant ep files problem". I have received mails from you and Greg that it's already applied to usb tree, so I have assumed that there is no need to include it to this patchset. I can resend this patch if it's needed. ^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [PATCH RESEND v7 2/2] usb: gadget: f_fs: virtual endpoint address mapping 2014-09-09 6:02 ` Robert Baldyga @ 2014-09-09 14:43 ` Felipe Balbi 0 siblings, 0 replies; 6+ messages in thread From: Felipe Balbi @ 2014-09-09 14:43 UTC (permalink / raw) To: Robert Baldyga; +Cc: balbi, gregkh, linux-usb, linux-kernel [-- Attachment #1: Type: text/plain, Size: 3235 bytes --] On Tue, Sep 09, 2014 at 08:02:16AM +0200, Robert Baldyga wrote: > On 09/08/2014 03:47 PM, Felipe Balbi wrote: > > Hi, > > > > On Mon, Sep 08, 2014 at 09:57:00AM +0200, Robert Baldyga wrote: > >> This patch introduces virtual endpoint address mapping. It separates > >> function logic form physical endpoint addresses making it more hardware > >> independent. > >> > >> Following modifications changes user space API, so to enable them user > >> have to switch on the FUNCTIONFS_VIRTUAL_ADDR flag in descriptors. > >> > >> Endpoints are now refered using virtual endpoint addresses chosen by > >> user in endpoint descpriptors. This applies to each context when endpoint > >> address can be used: > >> - when accessing endpoint files in FunctionFS filesystemi (in file name), > >> - in setup requests directed to specific endpoint (in wIndex field), > >> - in descriptors returned by FUNCTIONFS_ENDPOINT_DESC ioctl. > >> > >> In endpoint file names the endpoint address number is formatted as > >> double-digit hexadecimal value ("ep%02x") which has few advantages - > >> it is easy to parse, allows to easly recognize endpoint direction basing > >> on its name (IN endpoint number starts with digit 8, and OUT with 0) > >> which can be useful for debugging purpose, and it makes easier to introduce > >> further features allowing to use each endpoint number in both directions > >> to have more endpoints available for function if hardware supports this > >> (for example we could have ep01 which is endpoint 1 with OUT direction, > >> and ep81 which is endpoint 1 with IN direction). > >> > >> Physical endpoint address can be still obtained using ioctl named > >> FUNCTIONFS_ENDPOINT_REVMAP, but now it's not neccesary to handle > >> USB transactions properly. > >> > >> Signed-off-by: Robert Baldyga <r.baldyga@samsung.com> > >> Acked-by: Michal Nazarewicz <mina86@mina86.com> > > > > after this patch I get build errors: > > > > drivers/usb/gadget/function/f_fs.c: In function ‘ffs_epfiles_create’: > > drivers/usb/gadget/function/f_fs.c:1555:40: error: ‘struct ffs_data’ has no member named ‘eps_addrmap’ > > sprintf(epfiles->name, "ep%02x", ffs->eps_addrmap[i]); > > ^ > > drivers/usb/gadget/function/f_fs.c: In function ‘ffs_func_setup’: > > drivers/usb/gadget/function/f_fs.c:2900:19: error: ‘struct ffs_data’ has no member named ‘eps_addrmap’ > > ret = func->ffs->eps_addrmap[ret]; > > ^ > > make[3]: *** [drivers/usb/gadget/function/f_fs.o] Error 1 > > make[3]: *** Waiting for unfinished jobs.... > > make[2]: *** [drivers/usb/gadget/function] Error 2 > > make[1]: *** [drivers/usb/gadget] Error 2 > > make[1]: *** Waiting for unfinished jobs.... > > make: *** [drivers/usb/] Error 2 > > > > Array "eps_addrmap" was introduced in patch "usb: gadget: f_fs: fix the > redundant ep files problem". I have received mails from you and Greg > that it's already applied to usb tree, so I have assumed that there is > no need to include it to this patchset. I can resend this patch if it's > needed. No, no, I was didn't have -rc4 merged on my next. Now I do. -- balbi [-- Attachment #2: Digital signature --] [-- Type: application/pgp-signature, Size: 819 bytes --] ^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2014-09-09 14:43 UTC | newest] Thread overview: 6+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2014-09-08 7:56 [PATCH RESEND v7 0/2] usb: gadget: f_fs: userspace API fixes and improvements Robert Baldyga 2014-09-08 7:56 ` [PATCH RESEND v7 1/2] usb: gadget: f_fs: add ioctl returning ep descriptor Robert Baldyga 2014-09-08 7:57 ` [PATCH RESEND v7 2/2] usb: gadget: f_fs: virtual endpoint address mapping Robert Baldyga 2014-09-08 13:47 ` Felipe Balbi 2014-09-09 6:02 ` Robert Baldyga 2014-09-09 14:43 ` Felipe Balbi
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox