Re: [PATCH] Declare the file_operations struct as const

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sun, Aug 29, 2021 at 4:45 PM Bryan Brattlof <hello@xxxxxxxxxxxxxxxxx> wrote:
>
> Hi Krish
>
> On this day, August 29, 2021, thus sayeth Krish Jain:
> > On Sun, Aug 29, 2021 at 8:17 AM Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
> > >
> > > On Sun, Aug 29, 2021 at 04:13:43AM +0200, Krish Jain wrote:
> > > > On Sat, Aug 28, 2021 at 1:38 AM Bryan Brattlof <hello@xxxxxxxxxxxxxxxxx> wrote:
> > > > > As for your patch, I built the driver using:
> > > > >
> > > > >   $ make CCFLAGS=-Werror W=1 M=drivers/staging/android
> > > > >
> > > > > Which produced the following error:
> > > > >
> > > > >
> > > > > drivers/staging/android/ashmem.c: In function ‘ashmem_mmap’:
> > > > > drivers/staging/android/ashmem.c:380:2: warning: ‘static’ is not at beginning of declaration [-Wold-style-declaration]
> > > > >   380 |  const static struct file_operations vmfile_fops;
> > > > >       |  ^~~~~
> > > > > drivers/staging/android/ashmem.c:431:16: error: assignment of read-only variable ‘vmfile_fops’
> > > > >   431 |    vmfile_fops = *vmfile->f_op;
> > > > >       |                ^
> > > > > drivers/staging/android/ashmem.c:432:21: error: assignment of member ‘mmap’ in read-only object
> > > > >   432 |    vmfile_fops.mmap = ashmem_vmfile_mmap;
> > > > >       |                     ^
> > > > > drivers/staging/android/ashmem.c:433:34: error: assignment of member ‘get_unmapped_area’ in read-only object
> > > > >   433 |    vmfile_fops.get_unmapped_area =
> > > > >       |                                  ^
> > > > > make[1]: *** [scripts/Makefile.build:271: drivers/staging/android/ashmem.o] Error 1
> > > > > make: *** [Makefile:1851: drivers/staging/android] Error 2
> > > > >
> > > >
> > > > Hi, this seems very useful and I tried this myself just now. I don't
> > > > get any errors that you do though. When I hit enter I just get a new
> > > > shell prompt. What am I doing wrong? Probably a silly mistake. I ran
> > > > make CCFLAGS=-Werror M=drivers/staging/android/.
> > >
> > > Are you sure the file is being built at all?  You usually have to select
> > > the proper configuration option to enable that driver as well.
> >
> >
> > Hi, what option do you mean?  I already ran make allmodconfig and sudo
> > make modules_install install and then make   "CCFLAGS=-Werror W=1
> > M=drivers/staging/android/" and now I do get output but one line
> > "WARNING: Symbol version dump "Module.symvers" is missing. Modules may
> > not have dependencies or modversions. You may get many unresolved
> > symbol warnings." . Then I tried "make CCFLAGS=-Werror V=1
> > M=drivers/staging/android/" and that outputted the following:
> >
>
> Most of the answers you're asking for are going to get vague responses
> (if any) on the mailing lists. The idea being (and I agree with) that
> giving out the answers will steal your opportunity to explore and learn
> the material yourself.
>
> Yes, it would be faster if we told you the answer, but ultimately, we
> would be doing a disservice to you.
>
> Besides, more times than not we (me especially) don't have the answer.
>
> With that said, I will give a (generous) hint. :)
>

Hi. Do I have to build the kernel once before this works? Or can I
just build a module directly?



Best Regards


>
> It looks like your having trouble with Kconfig. Have a look at:
>
>    https://www.kernel.org/doc/html/latest/kbuild/makefiles.html
>
> and:
>
>    https://www.kernel.org/doc/html/latest/kbuild/modules.html
>
> Also, you shouldn't need to install anything if you're just testing
> whether a module builds. Especially when 'sudo' and 'install' are
> involved.
>
> >
> > test -e include/generated/autoconf.h -a -e include/config/auto.conf || ( \
> > echo >&2; \
> > echo >&2 "  ERROR: Kernel configuration is invalid."; \
> > echo >&2 "         include/generated/autoconf.h or
> > include/config/auto.conf are missing.";\
> > echo >&2 "         Run 'make oldconfig && make prepare' on kernel src
> > to fix it."; \
> >
>
> If I were you I would be asking:
>
> "What is a Kernel configuration file? and how did I corrupt mine?"
>
>    https://www.kernel.org/doc/html/latest/kbuild/kconfig.html
>
> >
> > echo >&2 ; \
> > /bin/false)
> > make -f ./scripts/Makefile.build obj=drivers/staging/android \
> > single-build= \
> > need-builtin=1 need-modorder=1
> > sh ./scripts/modules-check.sh drivers/staging/android/modules.order
> > make -f ./scripts/Makefile.modpost
> > WARNING: Symbol version dump "Module.symvers" is missing.
> >          Modules may not have dependencies or modversions.
> >          You may get many unresolved symbol warnings.
> > make -f ./scripts/Makefile.modfinal
> >
>
> "What is Module.symvers?"
>
>   If you're reading the links I gave, you should know this by now. :)
>   Check out chapter 6 in "Building External Modules"
>
> >
> > I followed this and ran make oldconfig && make prepare but all that is
> > outputted is again  "WARNING: Symbol version dump "Module.symvers" is
> > missing. Modules may not have dependencies or modversions. You may get
> > many unresolved symbol warnings." Then I just tried sudo make
> > modules_install install   again and what was outputted was:
> >
>
>   Again, Be *VERY* careful running commands you do not understand.
>   Especially when the words 'sudo' and 'install' are in the same
>   command.
>
> >
> > sed: can't read modules.order: No such file or directory
> > make[1]: *** [Makefile:1494: __modinst_pre] Error 2
> > make: *** [Makefile:351: __build_one_by_one] Error 2
> >
>
> "What does modules.order do?"
>
>   https://www.kernel.org/doc/html/latest/kbuild/kbuild.html
>
> >
> > Any ideas?  I've been stuck on debugging for hours to no avail. Please
> > enlighten me on where I am messing up.
> >
>
> All I can say is the kernels' documentation is the greatest thing to
> have when joining the community. The search bar is *fantastic*.
>
>   https://www.kernel.org/doc/html/latest/index.html
>
> Together with the collective history of ~30 years of developer's
> conversations on Linux Kernel Mailing List (LMKL).
>
>   https://lore.kernel.org/lkml/
>
> Along with our ability to see the full commit history at any time by:
>
>   $ git grep <regexp>
>
> In the end, all of the questions you have are in the documentation. I
> can't stress enough how appreciative we should be of contributers to
> Documentation/ and people like Jonathan Corbet who take the time to
> document how all of this stuff works in such a straight forward way.
>
> Withholding simple answers like this may seem unfair, but the juice is
> absolutely worth the squeeze the next time you run into another issue.
>
> And trust me when I say you *will* run into another issue. We wouldn't
> be here if there weren't constant issues to solve.
>
> We've all been where you're at. Even Greg, he may deny this, but I'm
> sure he has been.
>
> I'm excited to see what you learn.
> ~Bryan
>
> ***
>
> Just as an aside, for any other lurkers out there: This was the email
> that gave me confidence that I could join this community and that
> sending a patch, no matter how small, is welcome.
>
>   https://lore.kernel.org/lkml/Pine.LNX.4.58.0412201646220.4112@xxxxxxxxxxxxxxx/
>
> The old "grey-beards" here don't bite, no matter how much they and
> everyone else say they do.
>





[Index of Archives]     [Linux Driver Development]     [Linux Driver Backports]     [DMA Engine]     [Linux GPIO]     [Linux SPI]     [Video for Linux]     [Linux USB Devel]     [Linux Coverity]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]
  Powered by Linux