[PATCH 0/5] ia64: Fix compiler warnings

[PATCH 0/5] ia64: Fix compiler warnings

[Matt Fleming]

I routinely build ia64 kernels when merging EFI patches and for a
while now I've seen a bunch of warnings from GCC.

These patches silence those warnings, with the first patch fixing an
actual bug but the rest just making GCC happier.

NOTE: None of these patches have been runtime tested.

Matt Fleming (5):
  ia64/PCI: Fix incorrect PCI resource end address
  ia64/PCI: Remove unused 'addr' and fix build warning
  ia64: Reduce stack usage by iterating over nodemask
  ia64/traps: Silence GCC warning about uninitialised variable
  ia64/unaligned: Silence another GCC warning about an uninitialised
    variable

 arch/ia64/kernel/traps.c           |  1 +
 arch/ia64/kernel/unaligned.c       |  1 +
 arch/ia64/sn/kernel/io_acpi_init.c |  1 -
 arch/ia64/sn/kernel/io_init.c      |  4 ++--
 arch/ia64/sn/kernel/sn2/sn2_smp.c  | 35 +++++++++++++++++++++++------------
 5 files changed, 27 insertions(+), 15 deletions(-)

-- 
2.7.3

[PATCH 1/5] ia64/PCI: Fix incorrect PCI resource end address

[Matt Fleming]

commit f976721e826e ("ia64/PCI: Use ioremap() instead of open-coded
equivalent") introduced the following compiler warning,

  arch/ia64/sn/kernel/io_init.c: In function 'sn_io_slot_fixup':
  arch/ia64/sn/kernel/io_init.c:189:19: warning: 'addr' may be used uninitialized in this function [-Wmaybe-uninitialized]
     res->end = addr + size;
                   ^

'addr' is indeed uninitialised and the correct value to use is
res->start.

Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Bjorn Helgaas <helgaas@kernel.org>
Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
---
 arch/ia64/sn/kernel/io_init.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/ia64/sn/kernel/io_init.c b/arch/ia64/sn/kernel/io_init.c
index c15a41e2d1f2..d63809a6adfa 100644
--- a/arch/ia64/sn/kernel/io_init.c
+++ b/arch/ia64/sn/kernel/io_init.c
@@ -151,7 +151,7 @@ sn_io_slot_fixup(struct pci_dev *dev)
 {
 	int idx;
 	struct resource *res;
-	unsigned long addr, size;
+	unsigned long size;
 	struct pcidev_info *pcidev_info;
 	struct sn_irq_info *sn_irq_info;
 	int status;
@@ -186,7 +186,7 @@ sn_io_slot_fixup(struct pci_dev *dev)
 			continue;
 
 		res->start = pcidev_info->pdi_pio_mapped_addr[idx];
-		res->end = addr + size;
+		res->end = res->start + size;
 
 		/*
 		 * if it's already in the device structure, remove it before
-- 
2.7.3

[PATCH 2/5] ia64/PCI: Remove unused 'addr' and fix build warning

[Matt Fleming]

Ever since commit 240504adaf07 ("ia64/PCI: Keep CPU physical (not
virtual) addresses in shadow ROM resource") 'addr' has been unused,
resulting in the following compiler warning,

  arch/ia64/sn/kernel/io_acpi_init.c: In function 'sn_acpi_slot_fixup':
  arch/ia64/sn/kernel/io_acpi_init.c:429:16: warning: unused variable 'addr' [-Wunused-variable]
    void __iomem *addr;
                ^

Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Bjorn Helgaas <helgaas@kernel.org>
Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
---
 arch/ia64/sn/kernel/io_acpi_init.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/arch/ia64/sn/kernel/io_acpi_init.c b/arch/ia64/sn/kernel/io_acpi_init.c
index 231234c8d113..c31fe637b0b4 100644
--- a/arch/ia64/sn/kernel/io_acpi_init.c
+++ b/arch/ia64/sn/kernel/io_acpi_init.c
@@ -426,7 +426,6 @@ sn_acpi_get_pcidev_info(struct pci_dev *dev, struct pcidev_info **pcidev_info,
 void
 sn_acpi_slot_fixup(struct pci_dev *dev)
 {
-	void __iomem *addr;
 	struct pcidev_info *pcidev_info = NULL;
 	struct sn_irq_info *sn_irq_info = NULL;
 	struct resource *res;
-- 
2.7.3

[PATCH 3/5] ia64: Reduce stack usage by iterating over nodemask

[Matt Fleming]

GCC complains about sn2_global_tlb_purge() because of the large stack
required by the function,

  arch/ia64/sn/kernel/sn2/sn2_smp.c: In function 'sn2_global_tlb_purge':
  arch/ia64/sn/kernel/sn2/sn2_smp.c:319:1: warning: the frame size of 2176 bytes is larger than 2048 bytes [-Wframe-larger-than=]

2048 bytes of the stack are consumed by the node ID array 'nasids[]'.
But we don't actually need to put the ID array on the stack and can
use nodemask operations.

Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Bjorn Helgaas <helgaas@kernel.org>
Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
---
 arch/ia64/sn/kernel/sn2/sn2_smp.c | 35 +++++++++++++++++++++++------------
 1 file changed, 23 insertions(+), 12 deletions(-)

diff --git a/arch/ia64/sn/kernel/sn2/sn2_smp.c b/arch/ia64/sn/kernel/sn2/sn2_smp.c
index f9c8d9fc5939..c98dc965fe82 100644
--- a/arch/ia64/sn/kernel/sn2/sn2_smp.c
+++ b/arch/ia64/sn/kernel/sn2/sn2_smp.c
@@ -54,7 +54,7 @@ sn2_ptc_deadlock_recovery_core(volatile unsigned long *, unsigned long,
 			       volatile unsigned long *, unsigned long,
 			       volatile unsigned long *, unsigned long);
 void
-sn2_ptc_deadlock_recovery(short *, short, short, int,
+sn2_ptc_deadlock_recovery(nodemask_t, short, short, int,
 			  volatile unsigned long *, unsigned long,
 			  volatile unsigned long *, unsigned long);
 
@@ -169,7 +169,7 @@ sn2_global_tlb_purge(struct mm_struct *mm, unsigned long start,
 	int use_cpu_ptcga;
 	volatile unsigned long *ptc0, *ptc1;
 	unsigned long itc, itc2, flags, data0 = 0, data1 = 0, rr_value, old_rr = 0;
-	short nasids[MAX_NUMNODES], nix;
+	short nix;
 	nodemask_t nodes_flushed;
 	int active, max_active, deadlock, flush_opt = sn2_flush_opt;
 
@@ -218,9 +218,7 @@ sn2_global_tlb_purge(struct mm_struct *mm, unsigned long start,
 	}
 
 	itc = ia64_get_itc();
-	nix = 0;
-	for_each_node_mask(cnode, nodes_flushed)
-		nasids[nix++] = cnodeid_to_nasid(cnode);
+	nix = nodes_weight(nodes_flushed);
 
 	rr_value = (mm->context << 3) | REGION_NUMBER(start);
 
@@ -270,8 +268,10 @@ sn2_global_tlb_purge(struct mm_struct *mm, unsigned long start,
 			data0 = (data0 & ~SH2_PTC_ADDR_MASK) | (start & SH2_PTC_ADDR_MASK);
 		deadlock = 0;
 		active = 0;
-		for (ibegin = 0, i = 0; i < nix; i++) {
-			nasid = nasids[i];
+		ibegin = 0;
+		i = 0;
+		for_each_node_mask(cnode, nodes_flushed) {
+			nasid = cnodeid_to_nasid(cnode);
 			if (use_cpu_ptcga && unlikely(nasid = mynasid)) {
 				ia64_ptcga(start, nbits << 2);
 				ia64_srlz_i();
@@ -286,13 +286,14 @@ sn2_global_tlb_purge(struct mm_struct *mm, unsigned long start,
 				if ((deadlock = wait_piowc())) {
 					if (flush_opt = 1)
 						goto done;
-					sn2_ptc_deadlock_recovery(nasids, ibegin, i, mynasid, ptc0, data0, ptc1, data1);
+					sn2_ptc_deadlock_recovery(nodes_flushed, ibegin, i, mynasid, ptc0, data0, ptc1, data1);
 					if (reset_max_active_on_deadlock())
 						max_active = 1;
 				}
 				active = 0;
 				ibegin = i + 1;
 			}
+			i++;
 		}
 		start += (1UL << nbits);
 	} while (start < end);
@@ -327,11 +328,12 @@ done:
  */
 
 void
-sn2_ptc_deadlock_recovery(short *nasids, short ib, short ie, int mynasid,
+sn2_ptc_deadlock_recovery(nodemask_t nodes, short ib, short ie, int mynasid,
 			  volatile unsigned long *ptc0, unsigned long data0,
 			  volatile unsigned long *ptc1, unsigned long data1)
 {
 	short nasid, i;
+	int cnode;
 	unsigned long *piows, zeroval, n;
 
 	__this_cpu_inc(ptcstats.deadlocks);
@@ -339,17 +341,26 @@ sn2_ptc_deadlock_recovery(short *nasids, short ib, short ie, int mynasid,
 	piows = (unsigned long *) pda->pio_write_status_addr;
 	zeroval = pda->pio_write_status_val;
 
+	i = 0;
+	for_each_node_mask(cnode, nodes) {
+		if (i < ib)
+			goto next;
+
+		if (i > ie)
+			break;
 
-	for (i=ib; i <= ie; i++) {
-		nasid = nasids[i];
+		nasid = cnodeid_to_nasid(cnode);
 		if (local_node_uses_ptc_ga(is_shub1()) && nasid = mynasid)
-			continue;
+			goto next;
+
 		ptc0 = CHANGE_NASID(nasid, ptc0);
 		if (ptc1)
 			ptc1 = CHANGE_NASID(nasid, ptc1);
 
 		n = sn2_ptc_deadlock_recovery_core(ptc0, data0, ptc1, data1, piows, zeroval);
 		__this_cpu_add(ptcstats.deadlocks2, n);
+next:
+		i++;
 	}
 
 }
-- 
2.7.3

[PATCH 4/5] ia64/traps: Silence GCC warning about uninitialised variable

[Matt Fleming]

  arch/ia64/kernel/traps.c: In function 'ia64_fault':
  arch/ia64/kernel/traps.c:433:17: warning: 'siginfo.si_code' may be used uninitialized in this function [-Wmaybe-uninitialized]
    struct siginfo siginfo;

Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Bjorn Helgaas <helgaas@kernel.org>
Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
---
 arch/ia64/kernel/traps.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/ia64/kernel/traps.c b/arch/ia64/kernel/traps.c
index 6f7d4a4dcf24..77edd68c5161 100644
--- a/arch/ia64/kernel/traps.c
+++ b/arch/ia64/kernel/traps.c
@@ -548,6 +548,7 @@ ia64_fault (unsigned long vector, unsigned long isr, unsigned long ifa,
 			return;
 		}
 		switch (vector) {
+		      default:
 		      case 29:
 			siginfo.si_code = TRAP_HWBKPT;
 #ifdef CONFIG_ITANIUM
-- 
2.7.3

[PATCH 5/5] ia64/unaligned: Silence another GCC warning about an uninitialised variable

[Matt Fleming]

  arch/ia64/kernel/unaligned.c: In function 'ia64_handle_unaligned':
  arch/ia64/kernel/unaligned.c:1385:16: warning: 'u.l' may be used uninitialized in this function [-Wmaybe-uninitialized]
    opcode = (u.l >> IA64_OPCODE_SHIFT) & IA64_OPCODE_MASK;
                ^

Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
---
 arch/ia64/kernel/unaligned.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/ia64/kernel/unaligned.c b/arch/ia64/kernel/unaligned.c
index e7ae6088350a..7f0d31656b4d 100644
--- a/arch/ia64/kernel/unaligned.c
+++ b/arch/ia64/kernel/unaligned.c
@@ -1378,6 +1378,7 @@ ia64_handle_unaligned (unsigned long ifa, struct pt_regs *regs)
 	 * extract the instruction from the bundle given the slot number
 	 */
 	switch (ipsr->ri) {
+	      default:
 	      case 0: u.l = (bundle[0] >>  5); break;
 	      case 1: u.l = (bundle[0] >> 46) | (bundle[1] << 18); break;
 	      case 2: u.l = (bundle[1] >> 23); break;
-- 
2.7.3

RE: [PATCH 0/5] ia64: Fix compiler warnings

[Luck, Tony]

>  ia64/PCI: Fix incorrect PCI resource end address
>  ia64/PCI: Remove unused 'addr' and fix build warning
>  ia64: Reduce stack usage by iterating over nodemask
>  ia64/traps: Silence GCC warning about uninitialised variable
>  ia64/unaligned: Silence another GCC warning about an uninitialized variable

Applied all 5.  I was a little leery of the two that added "default:" cases
to switches ... but I don't see a cleaner way to unconfuse the compiler
without a more radical restructuring of the code.

Thanks

-Tony