Skip to contentSkip to conversation

or1k: Fix compilation hiccup

PATCH In review Eric Blake 7 participants 12 messages May 26, 2020 – Jun 8, 2020

lore
Share & tools
lore

Eric Blake

author
#0 lore raw

PATCH 1/1 or1k: Fix compilation hiccup

On my Fedora 32 machine, gcc 10.1.1 at -O2 (the default for a bare './configure') has a false-positive complaint:

CC or1k-softmmu/hw/openrisc/openrisc_sim.o
/home/eblake/qemu/hw/openrisc/openrisc_sim.c: In function ‘openrisc_sim_init’:
/home/eblake/qemu/hw/openrisc/openrisc_sim.c:87:42: error: ‘cpu_irqs[0]’ may be used uninitialized in this function [-Werror=maybe-uninitialized]

   87 |         sysbus_connect_irq(s, i, cpu_irqs[i][irq_pin]);
      |                                  ~~~~~~~~^~~

Initializing both pointers of cpu_irqs[] to NULL is sufficient to shut up the compiler, even though they are definitely assigned in openrisc_sim_init() prior to the inlined call to openrisc_sim_ompic_init() containing the line in question.

---
 hw/openrisc/openrisc_sim.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/openrisc/openrisc_sim.c +1 -1
index d08ce6181199..95011a8015b4 100644
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};
     qemu_irq serial_irq;
     int n;
     unsigned int smp_cpus = machine->smp.cpus;
Signature · 1 line
2.26.2
#1 lore raw
Patchew URL: https://patchew.org/QEMU/[email protected]/

Patchew URL: https://patchew.org/QEMU/[email protected]/

Hi,

This series seems to have some coding style problems. See output below for more information:

Message-id: [email protected]
Subject: [PATCH] or1k: Fix compilation hiccup
Type: series

=== TEST SCRIPT BEGIN ===
#!/bin/bash
git rev-parse base > /dev/null || exit 0
git config --local diff.renamelimit 0
git config --local diff.renames True
git config --local diff.algorithm histogram
./scripts/checkpatch.pl --mailback base..
=== TEST SCRIPT END ===

Switched to a new branch 'test'
d96d2fb or1k: Fix compilation hiccup

=== OUTPUT BEGIN ===
ERROR: spaces required around that '*' (ctx:WxV)
#33: FILE: hw/openrisc/openrisc_sim.c:132:

+    qemu_irq *cpu_irqs[2] = {};
              ^

total: 1 errors, 0 warnings, 8 lines checked

Commit d96d2fbbc5db (or1k: Fix compilation hiccup) has style problems, please review. If any of these errors are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.
=== OUTPUT END ===

Test command exited with code: 1

The full log is available at
http://patchew.org/logs/[email protected]/testing.checkpatch/?type=message.
---
Email generated automatically by Patchew [https://patchew.org/].
Please send your feedback to [email protected]

author
#2 lore raw
Quoted 13 lines · On 5/26/20 6:21 PM, [email protected] wrote:

Patchew URL: https://patchew.org/QEMU/[email protected]/

Hi,

This series seems to have some coding style problems. See output below for more information:

=== OUTPUT BEGIN ===
ERROR: spaces required around that '*' (ctx:WxV)
#33: FILE: hw/openrisc/openrisc_sim.c:132:

Quoted 8 lines
+    qemu_irq *cpu_irqs[2] = {};
               ^

total: 1 errors, 0 warnings, 8 lines checked

Commit d96d2fbbc5db (or1k: Fix compilation hiccup) has style problems, please review. If any of these errors are false positives report them to the maintainer, see
CHECKPATCH in MAINTAINERS.

False positive, due to 'qemu_irq' not following the normal naming conventions for typedefs.

Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org
Reviewed-by on v1 1/1
#3 lore raw
In reply to Eric Blake
Quoted 25 lines · Eric Blake

On my Fedora 32 machine, gcc 10.1.1 at -O2 (the default for a bare './configure') has a false-positive complaint:

CC or1k-softmmu/hw/openrisc/openrisc_sim.o
/home/eblake/qemu/hw/openrisc/openrisc_sim.c: In function ‘openrisc_sim_init’:
/home/eblake/qemu/hw/openrisc/openrisc_sim.c:87:42: error: ‘cpu_irqs[0]’ may be used uninitialized in this function [-Werror=maybe-uninitialized]

   87 |         sysbus_connect_irq(s, i, cpu_irqs[i][irq_pin]);
      |                                  ~~~~~~~~^~~

Initializing both pointers of cpu_irqs[] to NULL is sufficient to shut up the compiler, even though they are definitely assigned in openrisc_sim_init() prior to the inlined call to openrisc_sim_ompic_init() containing the line in question.

---
 hw/openrisc/openrisc_sim.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/openrisc/openrisc_sim.c +0 -0
index d08ce6181199..95011a8015b4 100644

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 135

Quoted 8 lines
hw/openrisc/openrisc_sim.c +1 -1
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};
     qemu_irq serial_irq;
     int n;
     unsigned int smp_cpus = machine->smp.cpus;
#4 lore raw
In reply to Eric Blake
Quoted 24 lines · Eric Blake

On my Fedora 32 machine, gcc 10.1.1 at -O2 (the default for a bare './configure') has a false-positive complaint:

CC or1k-softmmu/hw/openrisc/openrisc_sim.o
/home/eblake/qemu/hw/openrisc/openrisc_sim.c: In function ‘openrisc_sim_init’:
/home/eblake/qemu/hw/openrisc/openrisc_sim.c:87:42: error: ‘cpu_irqs[0]’ may be used uninitialized in this function [-Werror=maybe-uninitialized]

   87 |         sysbus_connect_irq(s, i, cpu_irqs[i][irq_pin]);
      |                                  ~~~~~~~~^~~

Initializing both pointers of cpu_irqs[] to NULL is sufficient to shut up the compiler, even though they are definitely assigned in openrisc_sim_init() prior to the inlined call to openrisc_sim_ompic_init() containing the line in question.

---
 hw/openrisc/openrisc_sim.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/openrisc/openrisc_sim.c +0 -0
index d08ce6181199..95011a8015b4 100644

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 132

Quoted 5 lines
hw/openrisc/openrisc_sim.c +1 -1
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};

Ah I now remembered why this warning sound familiar:
https://bugs.launchpad.net/qemu/+bug/1874073

Peter suggested a different fix, I tested it, and was expecting Martin Liska to post an updated patch.

Quoted 4 lines
hw/openrisc/openrisc_sim.c +0 -0
     qemu_irq serial_irq;
     int n;
     unsigned int smp_cpus = machine->smp.cpus;
#5 lore raw
In reply to Eric Blake

On 2020-05-26 at 20:51 CEST, Eric Blake wrote...

Quoted 24 lines

On my Fedora 32 machine, gcc 10.1.1 at -O2 (the default for a bare './configure') has a false-positive complaint:

CC or1k-softmmu/hw/openrisc/openrisc_sim.o
/home/eblake/qemu/hw/openrisc/openrisc_sim.c: In function ‘openrisc_sim_init’:
/home/eblake/qemu/hw/openrisc/openrisc_sim.c:87:42: error: ‘cpu_irqs[0]’ may be used uninitialized in this function [-Werror=maybe-uninitialized]

   87 |         sysbus_connect_irq(s, i, cpu_irqs[i][irq_pin]);
      |                                  ~~~~~~~~^~~

Initializing both pointers of cpu_irqs[] to NULL is sufficient to shut up the compiler, even though they are definitely assigned in openrisc_sim_init() prior to the inlined call to openrisc_sim_ompic_init() containing the line in question.

---
 hw/openrisc/openrisc_sim.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
hw/openrisc/openrisc_sim.c +0 -0
index d08ce6181199..95011a8015b4 100644

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 132

Quoted 5 lines
hw/openrisc/openrisc_sim.c +1 -1
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};

Why is the value [2] correct here? The loop that initializes loops over machine->smp.cpus. Is it always less than 2 on this machine?

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 135

Quoted 3 lines
hw/openrisc/openrisc_sim.c +0 -0
     qemu_irq serial_irq;
     int n;
     unsigned int smp_cpus = machine->smp.cpus;

--
Cheers,
Christophe de Dinechin (IRC c3d)

#6 lore raw
Quoted 14 lines · Christophe de Dinechin

On 2020-05-26 at 20:51 CEST, Eric Blake wrote...

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 132

Quoted 10 lines
hw/openrisc/openrisc_sim.c +1 -1
index d08ce6181199..95011a8015b4 100644
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};

Why is the value [2] correct here? The loop that initializes loops over machine->smp.cpus. Is it always less than 2 on this machine?

Yes: openrisc_sim_machine_init() sets mc->max_cpus = 2. My suggestion of adding an assert() is essentially telling the compiler that indeed smp_cpus must always be in the range [1,2], which we can tell but it can't.

thanks
-- PMM

#7 lore raw
Quoted 17 lines · Peter Maydell
Quoted 14 lines · Christophe de Dinechin

On 2020-05-26 at 20:51 CEST, Eric Blake wrote...

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 132

Quoted 10 lines
hw/openrisc/openrisc_sim.c +1 -1
index d08ce6181199..95011a8015b4 100644
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};

Why is the value [2] correct here? The loop that initializes loops over machine->smp.cpus. Is it always less than 2 on this machine?

Quoted 4 lines

Yes: openrisc_sim_machine_init() sets mc->max_cpus = 2. My suggestion of adding an assert() is essentially telling the compiler that indeed smp_cpus must always be in the range [1,2], which we can tell but it can't.

Do we have a proper patch for this on the list?

#8 lore raw
Quoted 19 lines · Markus Armbruster
Quoted 17 lines · Peter Maydell
Quoted 14 lines · Christophe de Dinechin

On 2020-05-26 at 20:51 CEST, Eric Blake wrote...

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 132

Quoted 10 lines
hw/openrisc/openrisc_sim.c +1 -1
index d08ce6181199..95011a8015b4 100644
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};

Why is the value [2] correct here? The loop that initializes loops over machine->smp.cpus. Is it always less than 2 on this machine?

Quoted 6 lines
Quoted 4 lines

Yes: openrisc_sim_machine_init() sets mc->max_cpus = 2. My suggestion of adding an assert() is essentially telling the compiler that indeed smp_cpus must always be in the range [1,2], which we can tell but it can't.

Do we have a proper patch for this on the list?

Apparently not.

Philippe did try Peter's suggestion, found it works, but then posted it only to Launchpad. Philippe, please post to the list, so we can finally get this fixed.

#9 lore raw
Quoted 26 lines · Markus Armbruster
Quoted 24 lines · Markus Armbruster
Quoted 21 lines · Peter Maydell
Quoted 14 lines · Christophe de Dinechin

On 2020-05-26 at 20:51 CEST, Eric Blake wrote...

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 132

Quoted 10 lines
hw/openrisc/openrisc_sim.c +1 -1
index d08ce6181199..95011a8015b4 100644
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};

Why is the value [2] correct here? The loop that initializes loops over machine->smp.cpus. Is it always less than 2 on this machine?

Yes: openrisc_sim_machine_init() sets mc->max_cpus = 2. My suggestion of adding an assert() is essentially telling the compiler that indeed smp_cpus must always be in the range [1,2], which we can tell but it can't.

Quoted 7 lines
Quoted 1 line

Do we have a proper patch for this on the list?

Apparently not.

Philippe did try Peter's suggestion, found it works, but then posted it only to Launchpad. Philippe, please post to the list, so we can finally get this fixed.

Sorry since it was Eric finding, I didn't understood I had to post it. Will do.

#10 lore raw
Quoted 29 lines · Philippe Mathieu-Daudé
Quoted 28 lines · Markus Armbruster
Quoted 25 lines · Markus Armbruster
Quoted 17 lines · Peter Maydell
Quoted 14 lines · Christophe de Dinechin

On 2020-05-26 at 20:51 CEST, Eric Blake wrote...

hw/openrisc/openrisc_sim.c static void openrisc_sim_init(MachineState *machine) · line 132

Quoted 10 lines
hw/openrisc/openrisc_sim.c +1 -1
index d08ce6181199..95011a8015b4 100644
@@ -129,7 +129,7 @@ static void openrisc_sim_init(MachineState *machine)
     const char *kernel_filename = machine->kernel_filename;
     OpenRISCCPU *cpu = NULL;
     MemoryRegion *ram;
-    qemu_irq *cpu_irqs[2];
+    qemu_irq *cpu_irqs[2] = {};

Why is the value [2] correct here? The loop that initializes loops over machine->smp.cpus. Is it always less than 2 on this machine?

Quoted 4 lines

Yes: openrisc_sim_machine_init() sets mc->max_cpus = 2. My suggestion of adding an assert() is essentially telling the compiler that indeed smp_cpus must always be in the range [1,2], which we can tell but it can't.

Do we have a proper patch for this on the list?

Quoted 8 lines
Quoted 5 lines

Apparently not.

Philippe did try Peter's suggestion, found it works, but then posted it only to Launchpad. Philippe, please post to the list, so we can finally get this fixed.

Sorry since it was Eric finding, I didn't understood I had to post it. Will do.

You didn't *have* to, but it'll help if you do :)

author
#11 lore raw
Quoted 9 lines · Markus Armbruster
Quoted 9 lines
Quoted 9 lines
Quoted 6 lines
Quoted 4 lines

Yes: openrisc_sim_machine_init() sets mc->max_cpus = 2. My suggestion of adding an assert() is essentially telling the compiler that indeed smp_cpus must always be in the range [1,2], which we can tell but it can't.

Do we have a proper patch for this on the list?

Apparently not.

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

Philippe did try Peter's suggestion, found it works, but then posted it only to Launchpad. Philippe, please post to the list, so we can finally get this fixed.

Sorry since it was Eric finding, I didn't understood I had to post it. Will do.

You didn't *have* to, but it'll help if you do :)

Now at https://lists.gnu.org/archive/html/qemu-devel/2020-06/msg01779.html

Signature · 3 lines
Eric Blake, Principal Software Engineer
Red Hat, Inc.           +1-919-301-3226
Virtualization:  qemu.org | libvirt.org