| Message ID | 20220716065412.539153-1-windhl@126.com (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
| Series | powerpc/powermac/feature: Fix refcount leak bug | expand |
| Context | Check | Description |
|---|---|---|
| snowpatch_ozlabs/github-powerpc_ppctests | success | Successfully ran 10 jobs. |
| snowpatch_ozlabs/github-powerpc_selftests | success | Successfully ran 10 jobs. |
| snowpatch_ozlabs/github-powerpc_sparse | success | Successfully ran 4 jobs. |
| snowpatch_ozlabs/github-powerpc_clang | success | Successfully ran 7 jobs. |
| snowpatch_ozlabs/github-powerpc_kernel_qemu | success | Successfully ran 23 jobs. |
On Sat, 16 Jul 2022 14:54:12 +0800, Liang He wrote: > In probe_one_macio(), we should call of_node_put() for the refernece > 'node' escaped out of the for_each_node_by_name() which has increased > its refcount. While the 'node' will finally escaped into a global > reference, we should still call of_node_put() in fail path which will > stop global reference creation. > > > [...] Applied to powerpc/next. [1/1] powerpc/powermac/feature: Fix refcount leak bug https://git.kernel.org/powerpc/c/d36337ce950ce8c57a8e4f61593f923cceaf0dd7 cheers
diff --git a/arch/powerpc/platforms/powermac/feature.c b/arch/powerpc/platforms/powermac/feature.c index 5cc958adba13..0382d20b5619 100644 --- a/arch/powerpc/platforms/powermac/feature.c +++ b/arch/powerpc/platforms/powermac/feature.c @@ -2632,31 +2632,31 @@ static void __init probe_one_macio(const char *name, const char *compat, int typ if (!macio_chips[i].of_node) break; if (macio_chips[i].of_node == node) - return; + goto out_put; } if (i >= MAX_MACIO_CHIPS) { printk(KERN_ERR "pmac_feature: Please increase MAX_MACIO_CHIPS !\n"); printk(KERN_ERR "pmac_feature: %pOF skipped\n", node); - return; + goto out_put; } addrp = of_get_pci_address(node, 0, &size, NULL); if (addrp == NULL) { printk(KERN_ERR "pmac_feature: %pOF: can't find base !\n", node); - return; + goto out_put; } addr = of_translate_address(node, addrp); if (addr == 0) { printk(KERN_ERR "pmac_feature: %pOF, can't translate base !\n", node); - return; + goto out_put; } base = ioremap(addr, (unsigned long)size); if (!base) { printk(KERN_ERR "pmac_feature: %pOF, can't map mac-io chip !\n", node); - return; + goto out_put; } if (type == macio_keylargo || type == macio_keylargo2) { const u32 *did = of_get_property(node, "device-id", NULL); @@ -2677,6 +2677,11 @@ static void __init probe_one_macio(const char *name, const char *compat, int typ macio_chips[i].rev = *revp; printk(KERN_INFO "Found a %s mac-io controller, rev: %d, mapped at 0x%p\n", macio_names[type], macio_chips[i].rev, macio_chips[i].base); + + return; + +out_put: + of_node_put(node); } static int __init
In probe_one_macio(), we should call of_node_put() for the refernece 'node' escaped out of the for_each_node_by_name() which has increased its refcount. While the 'node' will finally escaped into a global reference, we should still call of_node_put() in fail path which will stop global reference creation. Fixes: 51d3082fe6e5 ("[PATCH] powerpc: Unify udbg (#2)") Signed-off-by: Liang He <windhl@126.com> --- commit-51d3082fe6e5 first introduces the 'of_find_node_by_name' which involved the refcounting bug. arch/powerpc/platforms/powermac/feature.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-)