[Potential Spoof] Re: [Potential Spoof] Re: IPMID crashing

[Vijay Khemka]

I debugged through this and suspecting as ipmb crashing this. In old implementation, MAX_IPMI_BUFFER was defined as 64 which I have made change to 256 in our patch to cater our ipmb response which is more than 64 bytes. But new ipmid implementation doesn't use this parameter so I am not sure how size is being defined.

Vernon, Can you please let us know how the size of response is defined as I doubt that our response size is more than 64 bytes which is causing issues.

Regards
-Vijay

On 3/21/19, 3:34 PM, "openbmc on behalf of Vijay Khemka" <openbmc-bounces+vijaykhemka=fb.com at lists.ozlabs.org on behalf of vijaykhemka at fb.com> wrote:

    Vernon,
    It is getting corrupted while using libipmid. I see now following messages
    
    double free or corruption (!prev)
    Aborted
    
    
    Regards
    -Vijay
    
    On 3/21/19, 2:34 PM, "openbmc on behalf of Vijay Khemka" <openbmc-bounces+vijaykhemka=fb.com at lists.ozlabs.org on behalf of vijaykhemka at fb.com> wrote:
    
        I searched for this output string and found in libc-2.28.so and libc.a. I have no idea, what is it doing. I have executed it manually from command line and can see all oem functions it registers.
        
        On 3/21/19, 12:47 PM, "Vernon Mauery" <vernon.mauery at linux.intel.com> wrote:
        
            On 21-Mar-2019 07:38 PM, Vijay Khemka wrote:
            >Anyone is seeing ipmid crashing. I have checkedout openbmc today and seeing ipmid is crashing with following message
            >
            >realloc(): invalid next size
            
            What was it doing? What IPMI command was running when it crashed? What 
            provider libraries do you have it loading?
            
            --Vernon