Add struct flock to newlib fcntl.h.

Add struct flock to newlib fcntl.h.

The header was declaring F_GETFL and F_SETFL but not
declare the flock struct on which they operate.

R=mseaborn@chromium.org

Committed: https://src.chromium.org/viewvc/native_client?view=rev&revision=12464

[Mark Seaborn, 2013-11-08 22:32:35]

https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
File src/trusted/service_runtime/include/sys/fcntl.h (right):

https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
src/trusted/service_runtime/include/sys/fcntl.h:74: long  l_start;  /* relative
offset, in bytes */
The Linux fcntl(2) man page says that l_start and l_len have type off_t (which
is 64 bits on NaCl).

This seems worth fixing in case this definition gets used later.

https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
src/trusted/service_runtime/include/sys/fcntl.h:76: short l_pid;    /* returned
with F_GETLK */
The Linux fcntl(2) man page says that l_pid is of type pid_t, which I hope is 32
bits on NaCl.

[Sam Clegg, 2013-11-08 22:41:55]

On 2013/11/08 22:32:35, Mark Seaborn wrote:
>
https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
> File src/trusted/service_runtime/include/sys/fcntl.h (right):
> 
>
https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
> src/trusted/service_runtime/include/sys/fcntl.h:74: long  l_start;  /*
relative
> offset, in bytes */
> The Linux fcntl(2) man page says that l_start and l_len have type off_t (which
> is 64 bits on NaCl).
> 
> This seems worth fixing in case this definition gets used later.
> 
>
https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
> src/trusted/service_runtime/include/sys/fcntl.h:76: short l_pid;    /*
returned
> with F_GETLK */
> The Linux fcntl(2) man page says that l_pid is of type pid_t, which I hope is
32
> bits on NaCl.

Ok, I'll fix both these issues.

I was re-using these definitions from newlib's fcntl.h because that seems to be
how this file is mostly constructed.

[Sam Clegg, 2013-11-25 19:50:25]

https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
File src/trusted/service_runtime/include/sys/fcntl.h (right):

https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
src/trusted/service_runtime/include/sys/fcntl.h:74: long  l_start;  /* relative
offset, in bytes */
On 2013/11/08 22:32:35, Mark Seaborn wrote:
> The Linux fcntl(2) man page says that l_start and l_len have type off_t (which
> is 64 bits on NaCl).
> 
> This seems worth fixing in case this definition gets used later.

Done.

https://codereview.chromium.org/57663004/diff/30001/src/trusted/service_runti...
src/trusted/service_runtime/include/sys/fcntl.h:76: short l_pid;    /* returned
with F_GETLK */
On 2013/11/08 22:32:35, Mark Seaborn wrote:
> The Linux fcntl(2) man page says that l_pid is of type pid_t, which I hope is
32
> bits on NaCl.

Done.

[Mark Seaborn, 2013-11-26 18:55:57]

LGTM

[commit-bot: I haz the power, 2013-11-27 02:38:24]

CQ is trying da patch. Follow status at
https://nativeclient-status.appspot.com/cq/sbc@chromium.org/57663004/570001

[commit-bot: I haz the power, 2013-11-27 02:38:28]

Presubmit check for 57663004-570001 failed and returned exit status 1.

Running presubmit commit checks ...

** Presubmit Warnings **
/b/commit-queue/workdir/nacl/src/trusted/service_runtime/include/sys/fcntl.h:
cpplint
 
/b/commit-queue/workdir/nacl/src/trusted/service_runtime/include/sys/fcntl.h:72:
 Use int16/int64/etc, rather than the C type short  [runtime/int] [4]
/b/commit-queue/workdir/nacl/src/trusted/service_runtime/include/sys/fcntl.h:73:
 Use int16/int64/etc, rather than the C type short  [runtime/int] [4]
Done processing
/b/commit-queue/workdir/nacl/src/trusted/service_runtime/include/sys/fcntl.h
Total errors found: 2


Presubmit checks took 1.8s to calculate.

Was the presubmit check useful? Please send feedback & hate mail to
maruel@chromium.org!

[Sam Clegg, 2013-11-27 20:49:05]

Committed patchset #3 manually as r12464 (presubmit successful).

[Mark Seaborn, 2013-12-03 02:57:44]

This change seems to have caused llvm-test-suite on the PNaCl FYI bots to start
failing:

http://build.chromium.org/p/client.nacl.toolchain/builders/linux-pnacl-x86_64...

/mnt/data/b/build/slave/linux-pnacl-x86_64-tests-x86_32/build/native_client/toolchain/pnacl_linux_x86/usr/include/sys/fcntl.h:71:8:
error: redefinition of 'flock'
struct flock {
       ^
/mnt/data/b/build/slave/linux-pnacl-x86_64-tests-x86_32/build/native_client/pnacl/git/llvm-test-suite/MultiSource/Applications/sqlite3/sqlite3.c:23:8:
note: previous definition is here
struct flock {
       ^
1 warning and 1 error generated.

It looks like you need to revert this change:

http://git.chromium.org/gitweb/?p=native_client/pnacl-llvm-testsuite.git;a=bl...