On Thu, Apr 30, 2015 at 2:17 AM, Kweh, Hock Leong <hock.leong.kweh@xxxxxxxxx> wrote: >> -----Original Message----- >> From: Andy Lutomirski [mailto:luto@xxxxxxxxxxxxxx] >> Sent: Thursday, April 30, 2015 2:41 AM >> >> On Wed, Apr 29, 2015 at 4:23 AM, Kweh, Hock Leong >> <hock.leong.kweh@xxxxxxxxx> wrote: >> > >> > Dear communities, >> > >> > I agree with James. Due to different people may have different needs. >> > But from our side, we would just like to have a simple interface for >> > us to upload the efi capsule and perform update. We do not have any >> > use case or need to get info from QueryCapsuleUpdate(). Let me give a >> suggestion here: >> > please allow me to focus on deliver this simple loading interface and >> > upstream it. Then later whoever has the actual use case or needs on >> > the ioctl implementation, he or she could enhance base on this simple >> loading interface. >> > What do you guys think? >> > >> > Let me summarize the latest design idea: >> > - No longer leverage on firmware class but use misc device >> > - Do not use platform device but use device_create() >> > - User just need to perform "cat file.bin > /sys/.../capsule_loader" >> > in the shell >> >> If you do this, there's no need for the misc device. > > I do this so that in the future when someone want to implement the > Ioctl(), he or she can base on this and expand it. > >> >> > - File operation functions include: open(), read(), write() and >> > flush() >> > - Perform mutex lock in open() then release the mutex in flush() for >> avoiding >> > race condition / concurrent loading >> >> Make sure the mutex operation is killable, then, and maybe even >> interruptable. > > Okay. > >> >> > - Perform the capsule update and error return at flush() function >> > >> > Is there anything I missed? Any one still have concern with this idea? >> > Thanks for providing the ideas as well as the review. >> > >> >> If it works (and cat really does fail reliably), then it seems okay to me. >> >> However, since I like pulling increasing numbers of my hats, someone should >> verify that the common embedded cat implementations are also okay with >> this. For example, I haven't yet found any code in busybox's cat >> implementation that closes stdout. >> >> Given that the main targets of this (for now, at least) are embedded, this >> might be a problem. >> > > I think we shouldn't focus on the cat implementation for the close issue. > > My understanding about this action: > cat file.bin > /sys/..../capsule_loader > It is actually the ">" (IO redirection) who perform the open write & close > to this "/sys/..../capsule_loader" file note and not the "cat" do it. > So, I think your answer can be found at Shell source code. The shell opens capsule_loader and then execs the command. If you type: (cat file.bin) >/sys/.../captule_loader, then the command is a subshell and the subshell will close the file. (cat might also close it, but there will be two references.) If you type: cat file.bin >/sys/.../capsule_loader then the shell doesn't retain a reference to the file at all. --Andy -- To unsubscribe from this list: send the line "unsubscribe linux-efi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
