Re: Patch 8.2.3871
John Marriott wrote: > >> On 23-Dec-2021 05:20, Bram Moolenaar wrote: > >>> Patch 8.2.3871 > >>> Problem:List.c contains code for dict and blob. > >>> Solution: Refactor to put code where it belongs. (Yegappan Lakshmanan, > >>> closes #9386) > >>> Files: src/blob.c, src/dict.c, src/list.c, src/proto/blob.pro, > >>> src/proto/dict.pro, src/proto/list.pro, > >>> src/proto/strings.pro, > >>> src/strings.c, src/structs.h, > >>> src/testdir/test_filter_map.vim, > >>> src/testdir/test_listdict.vim, src/testdir/test_sort.vim > >>> > >>> > > Is that because of the "actually longer" thing? > Yeah I think so. > > Any idea how to avoid the error? > > > > The attached patch changes dictitem_alloc() to use the same mechanism as > dictitem_copy() just below it. > > What do you think? If this avoids the warning then it's fine. Thanks for the patch. -- % cat /usr/include/real_life.h void life(void); /// Bram Moolenaar -- b...@moolenaar.net -- http://www.Moolenaar.net \\\ /// \\\ \\\sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ /// \\\help me help AIDS victims -- http://ICCF-Holland.org/// -- -- You received this message from the "vim_dev" maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit http://www.vim.org/maillist.php --- You received this message because you are subscribed to the Google Groups "vim_dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to vim_dev+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/vim_dev/2021103038.E22421C0641%40moolenaar.net.
Re: Patch 8.2.3871
On 23-Dec-2021 06:19, Bram Moolenaar wrote:
John Marriott wrote:
On 23-Dec-2021 05:20, Bram Moolenaar wrote:
Patch 8.2.3871
Problem:List.c contains code for dict and blob.
Solution: Refactor to put code where it belongs. (Yegappan Lakshmanan,
closes #9386)
Files: src/blob.c, src/dict.c, src/list.c, src/proto/blob.pro,
src/proto/dict.pro, src/proto/list.pro, src/proto/strings.pro,
src/strings.c, src/structs.h, src/testdir/test_filter_map.vim,
src/testdir/test_listdict.vim, src/testdir/test_sort.vim
Is that because of the "actually longer" thing?
Yeah I think so.
Any idea how to avoid the error?
The attached patch changes dictitem_alloc() to use the same mechanism as
dictitem_copy() just below it.
What do you think?
Cheers
John
--
--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
---
You received this message because you are subscribed to the Google Groups "vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to vim_dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/vim_dev/f9082c33-d8c1-3ba8-91ac-3a32b80045e7%40internode.on.net.
--- dict.c.orig 2021-12-23 05:28:28.586848100 +1100
+++ dict.c 2021-12-23 07:12:53.587225300 +1100
@@ -222,11 +222,12 @@
dictitem_alloc(char_u *key)
{
dictitem_T *di;
+size_t len = STRLEN(key);
-di = alloc(offsetof(dictitem_T, di_key) + STRLEN(key) + 1);
+di = alloc(offsetof(dictitem_T, di_key) + len + 1);
if (di != NULL)
{
- STRCPY(di->di_key, key);
+ mch_memmove(di->di_key, key, len + 1);
di->di_flags = DI_FLAGS_ALLOC;
di->di_tv.v_lock = 0;
di->di_tv.v_type = VAR_UNKNOWN;
Re: Patch 8.2.3871
John Marriott wrote:
> On 23-Dec-2021 05:20, Bram Moolenaar wrote:
> > Patch 8.2.3871
> > Problem:List.c contains code for dict and blob.
> > Solution: Refactor to put code where it belongs. (Yegappan Lakshmanan,
> > closes #9386)
> > Files: src/blob.c, src/dict.c, src/list.c, src/proto/blob.pro,
> > src/proto/dict.pro, src/proto/list.pro, src/proto/strings.pro,
> > src/strings.c, src/structs.h, src/testdir/test_filter_map.vim,
> > src/testdir/test_listdict.vim, src/testdir/test_sort.vim
> >
> >
> After this patch, mingw64 (gcc 11.2.0) spits out this warning:
>
> gcc -c -I. -Iproto -DWIN32 -DWINVER=0x0603 -D_WIN32_WINNT=0x0603
> -DHAVE_PATHDEF -DFEAT_NORMAL -DHAVE_STDINT_H -D__USE_MINGW_ANSI_STDIO
> -pipe -march=native -Wall -O3 -fomit-frame-pointer -freg-struct-return
> -fpie -fPIE -DFEAT_GUI_MSWIN -DFEAT_CLIPBOARD dict.c -o gobjnative/dict.o
> In file included from dict.c:14:
> In function 'dictitem_alloc',
> inlined from 'dict_add_tv' at dict.c:491:12,
> inlined from 'dict_filter_map' at dict.c:1370:7:
> vim.h:1629:29: warning: 'strcpy' offset 0 from the object at ''
> is out of the bounds of referenced subobject 'di_key' with type
> 'char_u[1]' {aka 'unsigned char[1]'} at offset 0 [-Warray-bounds]
> 1629 | #define STRCPY(d, s) strcpy((char *)(d), (char *)(s))
> | ^~~~
> vim.h:1629:29: note: in definition of macro 'STRCPY'
> 1629 | #define STRCPY(d, s) strcpy((char *)(d), (char *)(s))
> | ^~
> In file included from vim.h:1878,
> from dict.c:14:
> dict.c: In function 'dict_filter_map':
> structs.h:1538:17: note: subobject 'di_key' declared here
> 1538 | char_u di_key[1]; // key (actually longer!)
> | ^~
>
Is that because of the "actually longer" thing?
Any idea how to avoid the error?
--
>From "know your smileys":
<<<:-{Worf (Never smiles anyways, so he's a bad smiley)
/// Bram Moolenaar -- b...@moolenaar.net -- http://www.Moolenaar.net \\\
/// \\\
\\\sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ ///
\\\help me help AIDS victims -- http://ICCF-Holland.org///
--
--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
---
You received this message because you are subscribed to the Google Groups
"vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to vim_dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/vim_dev/20211222191935.349F41C0641%40moolenaar.net.
Re: Patch 8.2.3871
Hi,
On Wed, Dec 22, 2021 at 10:46 AM John Marriott
wrote:
>
>
> On 23-Dec-2021 05:20, Bram Moolenaar wrote:
> > Patch 8.2.3871
> > Problem:List.c contains code for dict and blob.
> > Solution: Refactor to put code where it belongs. (Yegappan Lakshmanan,
> > closes #9386)
> > Files: src/blob.c, src/dict.c, src/list.c, src/proto/blob.pro,
> > src/proto/dict.pro, src/proto/list.pro, src/proto/strings.pro,
> > src/strings.c, src/structs.h, src/testdir/test_filter_map.vim,
> > src/testdir/test_listdict.vim, src/testdir/test_sort.vim
> >
> >
> After this patch, mingw64 (gcc 11.2.0) spits out this warning:
>
> gcc -c -I. -Iproto -DWIN32 -DWINVER=0x0603 -D_WIN32_WINNT=0x0603
> -DHAVE_PATHDEF -DFEAT_NORMAL -DHAVE_STDINT_H -D__USE_MINGW_ANSI_STDIO
> -pipe -march=native -Wall -O3 -fomit-frame-pointer -freg-struct-return
> -fpie -fPIE -DFEAT_GUI_MSWIN -DFEAT_CLIPBOARD dict.c -o gobjnative/dict.o
> In file included from dict.c:14:
> In function 'dictitem_alloc',
> inlined from 'dict_add_tv' at dict.c:491:12,
> inlined from 'dict_filter_map' at dict.c:1370:7:
> vim.h:1629:29: warning: 'strcpy' offset 0 from the object at ''
> is out of the bounds of referenced subobject 'di_key' with type
> 'char_u[1]' {aka 'unsigned char[1]'} at offset 0 [-Warray-bounds]
> 1629 | #define STRCPY(d, s)strcpy((char *)(d), (char *)(s))
>| ^~~~
> vim.h:1629:29: note: in definition of macro 'STRCPY'
> 1629 | #define STRCPY(d, s)strcpy((char *)(d), (char *)(s))
>| ^~
> In file included from vim.h:1878,
> from dict.c:14:
> dict.c: In function 'dict_filter_map':
> structs.h:1538:17: note: subobject 'di_key' declared here
> 1538 | char_u di_key[1]; // key (actually longer!)
>| ^~
>
>
This patch did not change the dictitem_T definition or modify the dict_add_tv()
or the dictitem_alloc() functions. It looks like GCC is complaining about using
the di_key buffer with size 1 (even though the size of the di_key array is
greater than 1). Interestingly GCC is not complaining about the use of this
function in other places.
- Yegappan
--
--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
---
You received this message because you are subscribed to the Google Groups
"vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to vim_dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/vim_dev/CAAW7x7kWkMoPEX8Cm%2B0%2Bqcam9QrEOGdgDxPxFyU%3DNRs3N%2BGRhA%40mail.gmail.com.
Re: Patch 8.2.3871
On 23-Dec-2021 05:20, Bram Moolenaar wrote:
Patch 8.2.3871
Problem:List.c contains code for dict and blob.
Solution: Refactor to put code where it belongs. (Yegappan Lakshmanan,
closes #9386)
Files: src/blob.c, src/dict.c, src/list.c, src/proto/blob.pro,
src/proto/dict.pro, src/proto/list.pro, src/proto/strings.pro,
src/strings.c, src/structs.h, src/testdir/test_filter_map.vim,
src/testdir/test_listdict.vim, src/testdir/test_sort.vim
After this patch, mingw64 (gcc 11.2.0) spits out this warning:
gcc -c -I. -Iproto -DWIN32 -DWINVER=0x0603 -D_WIN32_WINNT=0x0603
-DHAVE_PATHDEF -DFEAT_NORMAL -DHAVE_STDINT_H -D__USE_MINGW_ANSI_STDIO
-pipe -march=native -Wall -O3 -fomit-frame-pointer -freg-struct-return
-fpie -fPIE -DFEAT_GUI_MSWIN -DFEAT_CLIPBOARD dict.c -o gobjnative/dict.o
In file included from dict.c:14:
In function 'dictitem_alloc',
inlined from 'dict_add_tv' at dict.c:491:12,
inlined from 'dict_filter_map' at dict.c:1370:7:
vim.h:1629:29: warning: 'strcpy' offset 0 from the object at ''
is out of the bounds of referenced subobject 'di_key' with type
'char_u[1]' {aka 'unsigned char[1]'} at offset 0 [-Warray-bounds]
1629 | #define STRCPY(d, s) strcpy((char *)(d), (char *)(s))
| ^~~~
vim.h:1629:29: note: in definition of macro 'STRCPY'
1629 | #define STRCPY(d, s) strcpy((char *)(d), (char *)(s))
| ^~
In file included from vim.h:1878,
from dict.c:14:
dict.c: In function 'dict_filter_map':
structs.h:1538:17: note: subobject 'di_key' declared here
1538 | char_u di_key[1]; // key (actually longer!)
| ^~
Cheers
John
--
--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
---
You received this message because you are subscribed to the Google Groups "vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to vim_dev+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/vim_dev/c69a6da9-875d-32ca-11d8-07723c06d4ff%40internode.on.net.
Patch 8.2.3871
Patch 8.2.3871
Problem:List.c contains code for dict and blob.
Solution: Refactor to put code where it belongs. (Yegappan Lakshmanan,
closes #9386)
Files: src/blob.c, src/dict.c, src/list.c, src/proto/blob.pro,
src/proto/dict.pro, src/proto/list.pro, src/proto/strings.pro,
src/strings.c, src/structs.h, src/testdir/test_filter_map.vim,
src/testdir/test_listdict.vim, src/testdir/test_sort.vim
*** ../vim-8.2.3870/src/blob.c 2021-12-19 19:19:27.818424761 +
--- src/blob.c 2021-12-22 18:04:56.843931664 +
***
*** 410,487
}
/*
! * "remove({blob})" function
*/
void
blob_remove(typval_T *argvars, typval_T *rettv, char_u *arg_errmsg)
{
blob_T*b = argvars[0].vval.v_blob;
int error = FALSE;
long idx;
long end;
if (b != NULL && value_check_lock(b->bv_lock, arg_errmsg, TRUE))
return;
idx = (long)tv_get_number_chk([1], );
! if (!error)
{
! int len = blob_len(b);
! char_u *p;
! if (idx < 0)
! // count from the end
! idx = len + idx;
! if (idx < 0 || idx >= len)
! {
! semsg(_(e_blobidx), idx);
return;
}
! if (argvars[2].v_type == VAR_UNKNOWN)
{
! // Remove one item, return its value.
! p = (char_u *)b->bv_ga.ga_data;
! rettv->vval.v_number = (varnumber_T) *(p + idx);
! mch_memmove(p + idx, p + idx + 1, (size_t)len - idx - 1);
--b->bv_ga.ga_len;
}
! else
{
! blob_T *blob;
! // Remove range of items, return blob with values.
! end = (long)tv_get_number_chk([2], );
! if (error)
! return;
! if (end < 0)
! // count from the end
! end = len + end;
! if (end >= len || idx > end)
! {
! semsg(_(e_blobidx), end);
! return;
! }
! blob = blob_alloc();
! if (blob == NULL)
! return;
! blob->bv_ga.ga_len = end - idx + 1;
! if (ga_grow(>bv_ga, end - idx + 1) == FAIL)
! {
! vim_free(blob);
! return;
! }
! p = (char_u *)b->bv_ga.ga_data;
! mch_memmove((char_u *)blob->bv_ga.ga_data, p + idx,
! (size_t)(end - idx + 1));
! ++blob->bv_refcount;
! rettv->v_type = VAR_BLOB;
! rettv->vval.v_blob = blob;
!
! if (len - end - 1 > 0)
! mch_memmove(p + idx, p + end + 1, (size_t)(len - end - 1));
! b->bv_ga.ga_len -= end - idx + 1;
}
}
}
/*
--- 410,715
}
/*
! * "add(blob, item)" function
! */
! void
! blob_add(typval_T *argvars, typval_T *rettv)
! {
! blob_T*b = argvars[0].vval.v_blob;
! int error = FALSE;
! varnumber_T n;
!
! if (b == NULL)
! {
! if (in_vim9script())
! emsg(_(e_cannot_add_to_null_blob));
! return;
! }
!
! if (value_check_lock(b->bv_lock, (char_u *)N_("add() argument"), TRUE))
! return;
!
! n = tv_get_number_chk([1], );
! if (error)
! return;
!
! ga_append(>bv_ga, (int)n);
! copy_tv([0], rettv);
! }
!
! /*
! * "remove({blob}, {idx} [, {end}])" function
*/
void
blob_remove(typval_T *argvars, typval_T *rettv, char_u *arg_errmsg)
{
blob_T*b = argvars[0].vval.v_blob;
+ blob_T*newblob;
int error = FALSE;
long idx;
long end;
+ int len;
+ char_u*p;
if (b != NULL && value_check_lock(b->bv_lock, arg_errmsg, TRUE))
return;
idx = (long)tv_get_number_chk([1], );
! if (error)
! return;
!
! len = blob_len(b);
!
! if (idx < 0)
! // count from the end
! idx = len + idx;
! if (idx < 0 || idx >= len)
! {
! semsg(_(e_blobidx), idx);
! return;
! }
! if (argvars[2].v_type == VAR_UNKNOWN)
{
! // Remove one item, return its value.
! p = (char_u *)b->bv_ga.ga_data;
! rettv->vval.v_number = (varnumber_T) *(p + idx);
! mch_memmove(p + idx, p + idx + 1, (size_t)len - idx - 1);
! --b->bv_ga.ga_len;
! return;
! }
! // Remove range of items, return blob with values.
! end = (long)tv_get_number_chk([2], );
! if (error)
! return;
! if (end < 0)
! // count from the end
! end = len + end;
! if (end >= len || idx > end)
! {
! semsg(_(e_blobidx), end);
! return;
! }
! newblob = blob_alloc();
! if (newblob == NULL)
! return;
! newblo
