Fwd: Recent newlib/cygwin change breaking cris targets

Fwd: Recent newlib/cygwin change breaking cris targets

[Jeff Law]

-------- Forwarded Message --------
Subject: Re: Recent newlib/cygwin change breaking cris targets
Date: Sat, 17 Aug 2019 10:18:37 +0200
From: Corinna Vinschen <vinschen@redhat.com>
To: Jeff Law <law@redhat.com>

On Aug 16 12:35, Jeff Law wrote:
> Your most recent change to newlib-cygwin is breaking the cris-elf and
> crisv32-elf targets:
> 
> 
> 
> In file included from fstat.c:7:
> fstat.c: In function ‘_fstat’:
> fstat.c:29:22: error: ‘struct new_stat’ has no member named ‘st_atim’;
> did you mean ‘st_atime’?
>    29 |   buf->st_atime = ks.st_atime;
>       |                      ^~~~~~~~
> fstat.c:30:22: error: ‘struct new_stat’ has no member named ‘st_mtim’;
> did you mean ‘st_mtime’?
>    30 |   buf->st_mtime = ks.st_mtime;
>       |                      ^~~~~~~~
> fstat.c:31:22: error: ‘struct new_stat’ has no member named ‘st_ctim’;
> did you mean ‘st_ctime’?
>    31 |   buf->st_ctime = ks.st_ctime;
>       |                      ^~~~~~~~
> 
> I think you need to update libgloss/cris/gensyscalls.

Can you please send this to the newlib ML, ideally with patch?
My patch was the result of a discussion starting at

https://sourceware.org/ml/newlib/2019/msg00479.html


Thanks,
Corinna

Re: Recent newlib/cygwin change breaking cris targets

[Jeff Johnston]

I have posted a fix that should solve the problem.  I don't have a cris-elf
toolchain to test so if there are any issues,
please let me know.

-- Jeff J.

On Mon, Aug 19, 2019 at 11:09 AM Jeff Law <law@redhat.com> wrote:

>
>
>
> -------- Forwarded Message --------
> Subject: Re: Recent newlib/cygwin change breaking cris targets
> Date: Sat, 17 Aug 2019 10:18:37 +0200
> From: Corinna Vinschen <vinschen@redhat.com>
> To: Jeff Law <law@redhat.com>
>
> On Aug 16 12:35, Jeff Law wrote:
> > Your most recent change to newlib-cygwin is breaking the cris-elf and
> > crisv32-elf targets:
> >
> >
> >
> > In file included from fstat.c:7:
> > fstat.c: In function ‘_fstat’:
> > fstat.c:29:22: error: ‘struct new_stat’ has no member named ‘st_atim’;
> > did you mean ‘st_atime’?
> >    29 |   buf->st_atime = ks.st_atime;
> >       |                      ^~~~~~~~
> > fstat.c:30:22: error: ‘struct new_stat’ has no member named ‘st_mtim’;
> > did you mean ‘st_mtime’?
> >    30 |   buf->st_mtime = ks.st_mtime;
> >       |                      ^~~~~~~~
> > fstat.c:31:22: error: ‘struct new_stat’ has no member named ‘st_ctim’;
> > did you mean ‘st_ctime’?
> >    31 |   buf->st_ctime = ks.st_ctime;
> >       |                      ^~~~~~~~
> >
> > I think you need to update libgloss/cris/gensyscalls.
>
> Can you please send this to the newlib ML, ideally with patch?
> My patch was the result of a discussion starting at
>
> https://sourceware.org/ml/newlib/2019/msg00479.html
>
>
> Thanks,
> Corinna
>

Re: Recent newlib/cygwin change breaking cris targets

[Jeff Law]

On 8/19/19 4:05 PM, Jeff Johnston wrote:
> I have posted a fix that should solve the problem.  I don't have a
> cris-elf toolchain to test so if there are any issues,
> please let me know.
THanks.  My tester will pick it up ~24hrs after the commit across all
the targets it knows about.

Jeff

Re: Recent newlib/cygwin change breaking cris targets

[Jeff Law]

[-- Attachment #1: Type: text/plain, Size: 694 bytes --]

On 8/19/19 4:23 PM, Jeff Law wrote:
> On 8/19/19 4:05 PM, Jeff Johnston wrote:
>> I have posted a fix that should solve the problem.  I don't have a
>> cris-elf toolchain to test so if there are any issues,
>> please let me know.
> THanks.  My tester will pick it up ~24hrs after the commit across all
> the targets it knows about.
cris32-elf and crisv32-elf are now working.

visium has similar breakage which appears to be fixable via the attached
patch.  I'm not sure if I've got commit privs in newlib, so if it looks
good to you, I'd appreciate it if you could commit on my behalf.


That's the last fallout I see from those changes.  Everything else in my
tester is building OK.

jeff

[-- Attachment #2: P --]
[-- Type: text/plain, Size: 1233 bytes --]

diff --git a/libgloss/visium/io-gdb.c b/libgloss/visium/io-gdb.c
index 4f16b2f7b..e662226e3 100644
--- a/libgloss/visium/io-gdb.c
+++ b/libgloss/visium/io-gdb.c
@@ -94,9 +94,9 @@ __hosted_from_gdb_stat (const struct gdb_stat *gs,
   s->st_size = gs->st_size;
   s->st_blksize = gs->st_blksize;
   s->st_blocks = gs->st_blocks;
-  s->st_atime = gs->st_atime;
-  s->st_mtime = gs->st_mtime;
-  s->st_ctime = gs->st_ctime;
+  s->st_atime = gs->st_atim;
+  s->st_mtime = gs->st_mtim;
+  s->st_ctime = gs->st_ctim;
 }
 
 void
diff --git a/libgloss/visium/io.h b/libgloss/visium/io.h
index a1bb879fb..9cb76ea30 100644
--- a/libgloss/visium/io.h
+++ b/libgloss/visium/io.h
@@ -57,9 +57,9 @@ struct gdb_stat {
   uint64_t    st_size;    /* total size, in bytes */
   uint64_t    st_blksize; /* blocksize for filesystem I/O */
   uint64_t    st_blocks;  /* number of blocks allocated */
-  gdb_time_t  st_atime;   /* time of last access */
-  gdb_time_t  st_mtime;   /* time of last modification */
-  gdb_time_t  st_ctime;   /* time of last change */
+  gdb_time_t  st_atim;   /* time of last access */
+  gdb_time_t  st_mtim;   /* time of last modification */
+  gdb_time_t  st_ctim;   /* time of last change */
 };
 
 struct gdb_timeval {

Re: Recent newlib/cygwin change breaking cris targets

[Corinna Vinschen]

[-- Attachment #1: Type: text/plain, Size: 854 bytes --]

On Aug 20 09:26, Jeff Law wrote:
> On 8/19/19 4:23 PM, Jeff Law wrote:
> > On 8/19/19 4:05 PM, Jeff Johnston wrote:
> >> I have posted a fix that should solve the problem.  I don't have a
> >> cris-elf toolchain to test so if there are any issues,
> >> please let me know.
> > THanks.  My tester will pick it up ~24hrs after the commit across all
> > the targets it knows about.
> cris32-elf and crisv32-elf are now working.
> 
> visium has similar breakage which appears to be fixable via the attached
> patch.  I'm not sure if I've got commit privs in newlib, so if it looks
> good to you, I'd appreciate it if you could commit on my behalf.
> 
> 
> That's the last fallout I see from those changes.  Everything else in my
> tester is building OK.

Pushed.


Thanks,
Corinna

-- 
Corinna Vinschen
Cygwin Maintainer
Red Hat

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]