Add font_data_import() and update consoles to use it.

The implementation of font_data_import() is based on code from fbcon,
which supports overflow checks and crc32 checksums. Fbcon uses the crc32
checksum.

Newport_con now implements the same overflow checks as fbcon. As before,
this console does not support checksums, which are optional. Newport_con
can now also handle input font data with a vertical pitch other than 32
bytes. (The vertical pitch is the offset between two glyphs in the font
data.)

As an internal change, remove the const qualifier from the data field
if struct font_data. This allows font_data_import() to write the data
without type casting. For all users of the font data via font_data_t,
the stored data is still read only.

v3:
- fix typos

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Helge Deller <deller@gmx.de>

Add font_data_is_equal() and update consoles to use it.

Font data is equal if it has the same size and contains the same values
on all bytes. Only fbcon uses a crc32 checksum. If set in both operands
the checksums have to be equal.

The new helper also guarantees to not compare internal fonts against
fonts from user space. Internal fonts cannot be ref-counted, so making
them equal to user-space fonts with the same byte sequence results in
undefined behavior.

The test only compares data buffers. Their interpretation is up each
console. Therefore remove a width test in fbcon_set_font().

v3:
- rebase onto font_data_{get,put}()

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Helge Deller <deller@gmx.de>

Add font_data_get() and font_data_put(). Update consoles to use them
over REFCOUNT() and plain kfree().

Newly allocated font data starts with a reference count of 1. Loading
the font puts the previously loaded font. If the reference count reaches
zero, font_data_put() frees the font data.

The kernel stores a refcount of zero for internal font data. Invoking
font_data_get() and font_data_put() tests this internally and returns
success without further operation. From the caller's perspective,
getting and putting works the same for all font data.

Fbcon used the userfont flag distinguish between internal fonts and
fonts loaded by user space. Only the latter where refcounted. With the
new helper's automatic handling of internal font data, remove the
userfont flag from fbcon.

Newport_con uses a default font, FONT_DATA, until user space loads
custom font data. Remove all special cases for FONT_DATA, as the get
and put calls' read-only handling also covers this case.

v3:
- fix module linker error wrt font symbols (Nathan, Arnd)
- fix typos

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Helge Deller <deller@gmx.de>

Add font_data_size() and update consoles to use it.

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Helge Deller <deller@gmx.de>

Store font data as pointer to font_data_t instead of unsigned char.
Update consoles.

Pointers to font data refer to the raw data. There is a hidden header
before the data that contains additional state. Document the existing
layout and semantics of font_data_t.

The data field in struct vc_font can be used by any console. Therefore
it still points to plain data without the additional header. Fbcon sets
its value from struct fbcon_display.fontdata. Hence, update the size
test in fbcon_resize() to use struct fbcon_display.fontdata instead of
struct vc_font.data.

v3:
- fix typos (Helge)
v2:
- 'Font lookup' -> 'Font description' in <linux/font.h>

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Helge Deller <deller@gmx.de>

The character count in the font data is unused. The internal fonts also
do not set it. Remove FNTCHARCNT().

Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Helge Deller <deller@gmx.de>

This was done entirely with mindless brute force, using

    git grep -l '\<k[vmz]*alloc_objs*(.*, GFP_KERNEL)' |
        xargs sed -i 's/\(alloc_objs*(.*\), GFP_KERNEL)/\1)/'

to convert the new alloc_obj() users that had a simple GFP_KERNEL
argument to just drop that argument.

Note that due to the extreme simplicity of the scripting, any slightly
more complex cases spread over multiple lines would not be triggered:
they definitely exist, but this covers the vast bulk of the cases, and
the resulting diff is also then easier to check automatically.

For the same reason the 'flex' versions will be done as a separate
conversion.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

This is the result of running the Coccinelle script from
scripts/coccinelle/api/kmalloc_objs.cocci. The script is designed to
avoid scalar types (which need careful case-by-case checking), and
instead replace kmalloc-family calls that allocate struct or union
object instances:

Single allocations:	kmalloc(sizeof(TYPE), ...)
are replaced with:	kmalloc_obj(TYPE, ...)

Array allocations:	kmalloc_array(COUNT, sizeof(TYPE), ...)
are replaced with:	kmalloc_objs(TYPE, COUNT, ...)

Flex array allocations:	kmalloc(struct_size(PTR, FAM, COUNT), ...)
are replaced with:	kmalloc_flex(*PTR, FAM, COUNT, ...)

(where TYPE may also be *VAR)

The resulting allocations no longer return "void *", instead returning
"TYPE *".

Signed-off-by: Kees Cook <kees@kernel.org>

newport_show_logo() is only activated if CONFIG_LOGO_SGI_CLUT224 is
set (otherwise it is a NOP). This configuration item will be removed
in an upcoming change so instead, make it depend on LOGO_LINUX_CLUT224.

Signed-off-by: Vincent Mailhol <mailhol@kernel.org>
Signed-off-by: Helge Deller <deller@gmx.de>

This reverts commit 864f9963ec6b4b76d104d595ba28110b87158003.

The patch is wrong as it checks vc_origin against vc_screenbuf,
while in text mode it should compare against vga_vram_base.

As such it broke VGA text scrolling, which can be reproduced like this:
(1) boot a kernel that is configured to use text mode VGA-console
(2) type commands:  ls -l /usr/bin | less -S
(3) scroll up/down with cursor-down/up keys

Reported-by: Jari Ruusu <jariruusu@protonmail.com>
Cc: stable@vger.kernel.org
Cc: Yi Yang <yiyang13@huawei.com>
Cc: GONG Ruiqi <gongruiqi1@huawei.com>
Signed-off-by: Helge Deller <deller@gmx.de>