stdio-common: Add tests for formatted scanf input specifiers

Add a collection of tests for formatted scanf input specifiers covering
the b, d, i, o, u, x, and X integer conversions, the a, A, e, E, f, F,
g, and G floating-point conversions, and the [, c, and s character
conversions.  Also the hh, h, l, and ll length modifiers are covered
with the integer conversions as are the l and L length modifier with the
floating-point conversions.  The tests cover assignment suppressing and
the field width as well, verifying the number of assignments made, the
number of characters consumed and the value assigned.

Add the common test code here as well as test cases for scanf, and then
base Makefile infrastructure plus target-agnostic input data, for the
character conversions and the `char', `short', and `long long' integer
ones, signed and unsigned, with remaining input data and other functions
from the scanf family deferred to subsequent additions.

Keep input data disabled and referring to BZ #12701 for entries that are
currently incorrectly accepted as valid data, such as '0b' or '0x' with
the relevant integer conversions or sequences of an insufficient number
of characters with the c conversion.

Reviewed-by: Joseph Myers <josmyers@redhat.com>

59 files changed, 15485 insertions, 0 deletions

diff --git a/stdio-common/Makefile b/stdio-common/Makefile
index d3733d0c3d..9cf037e23c 100644
--- a/stdio-common/Makefile
+++ b/stdio-common/Makefile
@@ -51,6 +51,33 @@ nonfmt-xprintf-stems := \
 
 xprintf-stems := $(sort $(fmt-xprintf-stems) $(nonfmt-xprintf-stems))
 
+# List of markers for scanf family function tests.
+xscanf-funcs := s
+
+# List of data types and formats for individual per-conversion scanf tests.
+# Further conversions are provided by sysdeps.
+xscanf-kinds := char int real uint
+fmt-xscanf-char-convs := c
+fmts-xscanf-char := b c s
+fmt-xscanf-int-convs := char llong short
+fmts-xscanf-int := d i
+fmt-xscanf-real-convs :=
+fmts-xscanf-real := aa a ee e ff f gg g
+fmt-xscanf-uint-convs := uchar ullong ushort
+fmts-xscanf-uint := b o u xx x
+
+xscanf-convs = $(sort $(fmt-xscanf-char-convs) $(fmt-xscanf-int-convs) \
+		      $(fmt-xscanf-real-convs) $(fmt-xscanf-uint-convs))
+
+xscanf-srcs = $(foreach p,$(xscanf-funcs), \
+		$(foreach c,$(xscanf-convs),tst-scanf-format-$(p)-$(c)))
+
+xscanf-stems = $(sort $(foreach k,$(xscanf-kinds), \
+			$(foreach f,$(fmts-xscanf-$(k)), \
+			  $(foreach p,$(xscanf-funcs), \
+			    $(foreach c,$(fmt-xscanf-$(k)-convs), \
+			      tst-scanf-format-$(p)-$(c)-$(f))))))
+
 headers := \
   bits/printf-ldbl.h \
   bits/stdio_lim.h \
@@ -367,6 +394,7 @@ tests-internal = \
 
 test-srcs = \
   $(xprintf-srcs) \
+  $(xscanf-srcs) \
   tst-fwrite-bz29459 \
   tst-printf \
   tst-printfsz-islongdouble \
@@ -376,6 +404,7 @@ test-srcs = \
 ifeq ($(run-built-tests),yes)
 tests-special += \
   $(foreach f,$(xprintf-stems),$(objpfx)$(f).out) \
+  $(foreach f,$(xscanf-stems),$(objpfx)$(f).out) \
   $(objpfx)tst-fwrite-bz29459.out \
   $(objpfx)tst-printf.out \
   $(objpfx)tst-printfsz-islongdouble.out \
@@ -387,6 +416,7 @@ ifeq (yes,$(build-shared))
 ifneq ($(PERL),no)
 tests-special += \
   $(foreach f,$(xprintf-stems),$(objpfx)$(f)-mem.out) \
+  $(foreach f,$(xscanf-stems),$(objpfx)$(f)-mem.out) \
   $(objpfx)tst-freopen2-mem.out \
   $(objpfx)tst-freopen3-mem.out \
   $(objpfx)tst-freopen4-mem.out \
@@ -409,6 +439,8 @@ tests-special += \
 generated += \
   $(foreach f,$(xprintf-stems),$(f)-mem.out) \
   $(foreach f,$(xprintf-stems),$(f).mtrace) \
+  $(foreach f,$(xscanf-stems),$(f)-mem.out) \
+  $(foreach f,$(xscanf-stems),$(f).mtrace) \
   tst-freopen2-mem.out \
   tst-freopen2.mtrace \
   tst-freopen3-mem.out \
@@ -601,6 +633,26 @@ $(objpfx)tst-printf-format-%.out: \
 	$(make-tst-printf-format-out) > $@; \
 	$(evaluate-test)
 
+# With scanf format tests reuse the same executable for all the inputs
+# of the same data type that use a different conversion specifier each.
+define tst-scanf-format-rule
+$$(objpfx)tst-scanf-format-$(1)-$(2)-$(3).out: \
+	tst-scanf-format-$(2)-$(3).input $$(objpfx)tst-scanf-format-$(1)-$(2)
+	$$(make-tst-scanf-format-out) > $$@ < $$<; \
+	$$(evaluate-test)
+endef
+make-tst-scanf-format-out = \
+  $(test-wrapper-env) \
+  $(run-program-env) \
+  MALLOC_TRACE=$(@:.out=.mtrace) \
+  LD_PRELOAD=$(common-objpfx)malloc/libc_malloc_debug.so \
+  $(host-test-program-cmd)
+$(foreach k,$(xscanf-kinds), \
+  $(foreach f,$(fmts-xscanf-$(k)), \
+    $(foreach p,$(xscanf-funcs), \
+      $(foreach c,$(fmt-xscanf-$(k)-convs), \
+	$(eval $(call tst-scanf-format-rule,$(p),$(c),$(f)))))))
+
 $(objpfx)tst-printfsz-islongdouble.out: \
   tst-printfsz-islongdouble.sh $(objpfx)tst-printfsz-islongdouble
 	$(SHELL) $^ '$(test-program-prefix)' $@; \
@@ -713,5 +765,13 @@ $(objpfx)tst-setvbuf1-cmp.out: tst-setvbuf1.expect $(objpfx)tst-setvbuf1.out
 $(objpfx)tst-printf-round: $(libm)
 $(objpfx)tst-scanf-round: $(libm)
 
+# With scanf format tests real conversions depend on libm.
+define tst-scanf-format-libm-dep
+$$(objpfx)tst-scanf-format-$(1)-$(2): $$(libm)
+endef
+$(foreach p,$(xscanf-funcs), \
+  $(foreach c,$(fmt-xscanf-real-convs), \
+    $(eval $(call tst-scanf-format-libm-dep,$(p),$(c)))))
+
 $(objpfx)tst-freopen7: $(shared-thread-library)
 $(objpfx)tst-freopen64-7: $(shared-thread-library)
diff --git a/stdio-common/tst-scanf-format-c-b.input b/stdio-common/tst-scanf-format-c-b.input
new file mode 100644
index 0000000000..52b2dc1a4c
--- /dev/null
+++ b/stdio-common/tst-scanf-format-c-b.input
@@ -0,0 +1,90 @@
+%[bcdefghijklmnopqrstuvw]::0:-1:
+%*[bcdefghijklmnopqrstuvw]::0:-1:
+%[bcdefghijklmnopqrstuvw]:The:0:-1:
+%*[bcdefghijklmnopqrstuvw]:The:0:-1:
+%[bcdefghijklmnopqrstuvw]:quick:1:5:quick:
+%*[bcdefghijklmnopqrstuvw]:quick:0:5:
+%[bcdefghijklmnopqrstuvw]:brown fox:1:5:brown:
+%*[bcdefghijklmnopqrstuvw]:brown fox:0:5:
+%[bcdefghijklmnopqrstuvw]:jumps over the lazy dog:1:5:jumps:
+%*[bcdefghijklmnopqrstuvw]:jumps over the lazy dog:0:5:
+%[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%*[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%[ Tabcdefghijklmnopqrstuvwxyz]:The:1:3:The:
+%*[ Tabcdefghijklmnopqrstuvwxyz]:The:0:3:
+%[ Tabcdefghijklmnopqrstuvwxyz]:quick:1:5:quick:
+%*[ Tabcdefghijklmnopqrstuvwxyz]:quick:0:5:
+%[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:1:9:brown fox:
+%*[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:0:9:
+%[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:1:23:jumps over the lazy dog:
+%*[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:0:23:
+%1[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%*1[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%1[ Tabcdefghijklmnopqrstuvwxyz]:The:1:1:T:
+%*1[ Tabcdefghijklmnopqrstuvwxyz]:The:0:1:
+%1[ Tabcdefghijklmnopqrstuvwxyz]:quick:1:1:q:
+%*1[ Tabcdefghijklmnopqrstuvwxyz]:quick:0:1:
+%1[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:1:1:b:
+%*1[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:0:1:
+%1[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:1:1:j:
+%*1[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:0:1:
+%2[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%*2[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%2[ Tabcdefghijklmnopqrstuvwxyz]:The:1:2:Th:
+%*2[ Tabcdefghijklmnopqrstuvwxyz]:The:0:2:
+%2[ Tabcdefghijklmnopqrstuvwxyz]:quick:1:2:qu:
+%*2[ Tabcdefghijklmnopqrstuvwxyz]:quick:0:2:
+%2[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:1:2:br:
+%*2[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:0:2:
+%2[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:1:2:ju:
+%*2[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:0:2:
+%5[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%*5[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%5[ Tabcdefghijklmnopqrstuvwxyz]:The:1:3:The:
+%*5[ Tabcdefghijklmnopqrstuvwxyz]:The:0:3:
+%5[ Tabcdefghijklmnopqrstuvwxyz]:quick:1:5:quick:
+%*5[ Tabcdefghijklmnopqrstuvwxyz]:quick:0:5:
+%5[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:1:5:brown:
+%*5[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:0:5:
+%5[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:1:5:jumps:
+%*5[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:0:5:
+%10[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%*10[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%10[ Tabcdefghijklmnopqrstuvwxyz]:The:1:3:The:
+%*10[ Tabcdefghijklmnopqrstuvwxyz]:The:0:3:
+%10[ Tabcdefghijklmnopqrstuvwxyz]:quick:1:5:quick:
+%*10[ Tabcdefghijklmnopqrstuvwxyz]:quick:0:5:
+%10[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:1:9:brown fox:
+%*10[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:0:9:
+%10[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:1:10:jumps over:
+%*10[ Tabcdefghijklmnopqrstuvwxyz]:jumps over the lazy dog:0:10:
+%25[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%*25[ Tabcdefghijklmnopqrstuvwxyz]::0:-1:
+%25[ Tabcdefghijklmnopqrstuvwxyz]:The:1:3:The:
+%*25[ Tabcdefghijklmnopqrstuvwxyz]:The:0:3:
+%25[ Tabcdefghijklmnopqrstuvwxyz]:quick:1:5:quick:
+%*25[ Tabcdefghijklmnopqrstuvwxyz]:quick:0:5:
+%25[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:1:9:brown fox:
+%*25[ Tabcdefghijklmnopqrstuvwxyz]:brown fox:0:9: