linux-kernel - [PATCH v2 6/8] vt: add ucs_get

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20250507141535.40655-7-nico@fluxnic.net>
Date: Wed,  7 May 2025 10:13:21 -0400
From: Nicolas Pitre <nico@...xnic.net>
To: Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	Jiri Slaby <jirislaby@...nel.org>
Cc: Nicolas Pitre <npitre@...libre.com>,
	linux-serial@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: [PATCH v2 6/8] vt: add ucs_get_fallback()

From: Nicolas Pitre <npitre@...libre.com>

This is the code querying the newly introduced tables.

Signed-off-by: Nicolas Pitre <npitre@...libre.com>
---
 drivers/tty/vt/Makefile    |  3 +-
 drivers/tty/vt/ucs.c       | 84 ++++++++++++++++++++++++++++++++++++++
 include/linux/consolemap.h |  6 +++
 3 files changed, 92 insertions(+), 1 deletion(-)

diff --git a/drivers/tty/vt/Makefile b/drivers/tty/vt/Makefile
index 509362a3e11e..ae746dcdeec8 100644
--- a/drivers/tty/vt/Makefile
+++ b/drivers/tty/vt/Makefile
@@ -36,7 +36,8 @@ $(obj)/defkeymap.c: $(obj)/%.c: $(src)/%.map
 
 endif
 
-$(obj)/ucs.o: $(src)/ucs.c $(obj)/ucs_width_table.h $(obj)/ucs_recompose_table.h
+$(obj)/ucs.o:	$(src)/ucs.c $(obj)/ucs_width_table.h \
+		$(obj)/ucs_recompose_table.h $(obj)/ucs_fallback_table.h
 
 # You may uncomment one of those to have the UCS tables be regenerated
 # during the build process. By default the _shipped versions are used.
diff --git a/drivers/tty/vt/ucs.c b/drivers/tty/vt/ucs.c
index b0b23830170d..6c15c5deda5b 100644
--- a/drivers/tty/vt/ucs.c
+++ b/drivers/tty/vt/ucs.c
@@ -157,3 +157,87 @@ u32 ucs_recompose(u32 base, u32 mark)
 
 	return result ? result->recomposed : 0;
 }
+
+/*
+ * The fallback table structures implement a 2-level lookup.
+ */
+
+struct ucs_page_desc {
+	u8 page;	/* Page index (high byte of code points) */
+	u8 count;	/* Number of entries in this page */
+	u16 start;	/* Start index in entries array */
+};
+
+struct ucs_page_entry {
+	u8 offset;	/* Offset within page (0-255) */
+	u8 fallback;	/* Fallback character or range start marker */
+};
+
+#include "ucs_fallback_table.h"
+
+static int ucs_page_desc_cmp(const void *key, const void *element)
+{
+	u8 page = *(u8 *)key;
+	const struct ucs_page_desc *entry = element;
+
+	if (page < entry->page)
+		return -1;
+	if (page > entry->page)
+		return 1;
+	return 0;
+}
+
+static int ucs_page_entry_cmp(const void *key, const void *element)
+{
+	u8 offset = *(u8 *)key;
+	const struct ucs_page_entry *entry = element;
+
+	if (offset < entry->offset)
+		return -1;
+	if (entry->fallback == UCS_PAGE_ENTRY_RANGE_MARKER) {
+		if (offset > entry[1].offset)
+			return 1;
+	} else {
+		if (offset > entry->offset)
+			return 1;
+	}
+	return 0;
+}
+
+/**
+ * ucs_get_fallback() - Get a substitution for the provided Unicode character
+ * @base: Base Unicode code point (UCS-4)
+ *
+ * Get a simpler fallback character for the provided Unicode character.
+ * This is used for terminal display when corresponding glyph is unavailable.
+ * The substitution may not be as good as the actual glyph for the original
+ * character but still way more helpful than a squared question mark.
+ *
+ * Return: Fallback Unicode code point, or 0 if none is available
+ */
+u32 ucs_get_fallback(u32 cp)
+{
+	const struct ucs_page_desc *page;
+	const struct ucs_page_entry *entry;
+	u8 page_idx = cp >> 8, offset = cp;
+
+	if (!UCS_IS_BMP(cp))
+		return 0;
+
+	page = __inline_bsearch(&page_idx, ucs_fallback_pages,
+				ARRAY_SIZE(ucs_fallback_pages),
+				sizeof(*ucs_fallback_pages),
+				ucs_page_desc_cmp);
+	if (!page)
+		return 0;
+
+	entry = __inline_bsearch(&offset, ucs_fallback_entries + page->start,
+				 page->count, sizeof(*ucs_fallback_entries),
+				 ucs_page_entry_cmp);
+	if (!entry)
+		return 0;
+
+	if (entry->fallback == UCS_PAGE_ENTRY_RANGE_MARKER)
+		entry++;
+	return entry->fallback;
+}
diff --git a/include/linux/consolemap.h b/include/linux/consolemap.h
index 8167494229db..6180b803795c 100644
--- a/include/linux/consolemap.h
+++ b/include/linux/consolemap.h
@@ -31,6 +31,7 @@ void console_map_init(void);
 bool ucs_is_double_width(uint32_t cp);
 bool ucs_is_zero_width(uint32_t cp);
 u32 ucs_recompose(u32 base, u32 mark);
+u32 ucs_get_fallback(u32 cp);
 #else
 static inline u16 inverse_translate(const struct vc_data *conp, u16 glyph,
 		bool use_unicode)
@@ -75,6 +76,11 @@ static inline u32 ucs_recompose(u32 base, u32 mark)
 {
 	return 0;
 }
+
+static inline u32 ucs_get_fallback(u32 cp)
+{
+	return 0;
+}
 #endif /* CONFIG_CONSOLE_TRANSLATIONS */
 
 #endif /* __LINUX_CONSOLEMAP_H__ */
-- 
2.49.0