Re: [dpdk-dev] [PATCH] mmap(2) returns MAP_FAILED, not NULL, on failure

[Matej Vido]

On 02.12.2017 01:54, Ferruh Yigit wrote:

On 11/30/2017 10:51 PM, Michael McConville wrote:

Signed-off-by: Michael McConville <mm...@mykolab.com>
---
lib/librte_eal/bsdapp/eal/eal_memory.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/librte_eal/bsdapp/eal/eal_memory.c 
b/lib/librte_eal/bsdapp/eal/eal_memory.c
index 6ba058578..2c8a4b592 100644
--- a/lib/librte_eal/bsdapp/eal/eal_memory.c
+++ b/lib/librte_eal/bsdapp/eal/eal_memory.c
@@ -155,7 +155,7 @@ rte_eal_hugepage_attach(void)
        /* Map the shared hugepage_info into the process address spaces */
        hpi = mmap(NULL, sizeof(struct hugepage_info), PROT_READ, MAP_PRIVATE,
                        fd_hugepage_info, 0);
-       if (hpi == NULL) {
+       if (hpi == MAP_FAILED) {

Hi Matej,

Can you fix same thing in szedata2 PMD please [1] ?

[1]
http://dpdk.org/browse/dpdk/tree/drivers/net/szedata2/rte_eth_szedata2.c?h=v17.11#n1556

Hi Ferruh,

I will send a patch.

Thanks,
Matej

                RTE_LOG(ERR, EAL, "Could not mmap %s\n", 
eal_hugepage_info_path());
                goto error;
        }