linux-kernel - Re: [PATCH v19 030/130] KVM: TDX: Add helper functions to print TDX SEAMCALL error

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Thu, 21 Mar 2024 16:52:14 -0700
From: Isaku Yamahata <isaku.yamahata@...el.com>
To: "Huang, Kai" <kai.huang@...el.com>
Cc: Isaku Yamahata <isaku.yamahata@...el.com>, kvm@...r.kernel.org,
	linux-kernel@...r.kernel.org, isaku.yamahata@...il.com,
	Paolo Bonzini <pbonzini@...hat.com>, erdemaktas@...gle.com,
	Sean Christopherson <seanjc@...gle.com>,
	Sagi Shahar <sagis@...gle.com>, chen.bo@...el.com,
	hang.yuan@...el.com, tina.zhang@...el.com,
	Binbin Wu <binbin.wu@...ux.intel.com>,
	Yuan Yao <yuan.yao@...el.com>, isaku.yamahata@...ux.intel.com
Subject: Re: [PATCH v19 030/130] KVM: TDX: Add helper functions to print TDX
 SEAMCALL error

On Thu, Mar 21, 2024 at 12:09:57PM +1300,
"Huang, Kai" <kai.huang@...el.com> wrote:

> > Does it make sense?
> > 
> > void pr_tdx_error(u64 op, u64 error_code)
> > {
> >          pr_err_ratelimited("SEAMCALL (0x%016llx) failed: 0x%016llx\n",
> >                             op, error_code);
> > }
> 
> Should we also have a _ret version?
> 
> void pr_seamcall_err(u64 op, u64 err)
> {
> 	/* A comment to explain why using the _ratelimited() version? */

Because KVM can hit successive seamcall erorrs e.g. during desutructing TD,
(it's unintentional sometimes), ratelimited version is preferred as safe guard.
For example, SEAMCALL on all or some LPs (TDH_MNG_KEY_FREEID) can fail at the
same time.  And the number of LPs can be hundreds.


> 	pr_err_ratelimited(...);
> }
> 
> void pr_seamcall_err_ret(u64 op, u64 err, struct tdx_module_args *arg)
> {
> 	pr_err_seamcall(op, err);
> 	
> 	pr_err_ratelimited(...);
> }
> 
> (Hmm... if you look at the tdx.c in TDX host, there's similar code there,
> and again, it was a little bit annoying when I did that..)
> 
> Again, if we just use seamcall_ret() for ALL SEAMCALLs except VP.ENTER, we
> can simply have one..

What about this?

void pr_seamcall_err_ret(u64 op, u64 err, struct tdx_module_args *arg)
{
        pr_err_ratelimited("SEAMCALL (0x%016llx) failed: 0x%016llx\n",
                           op, error_code);
        if (arg)	
        	pr_err_ratelimited(...);
}



> > void pr_tdx_sept_error(u64 op, u64 error_code, const union tdx_sept_entry *entry,
> > 		       const union tdx_sept_level_state *level_state)
> > {
> > #define MSG \
> >          "SEAMCALL (0x%016llx) failed: 0x%016llx entry 0x%016llx level_state 0x%016llx\n"
> >          pr_err_ratelimited(MSG, op, error_code, entry->raw, level_state->raw);
> > }
> 
> A higher-level wrapper to print SEPT error is fine to me, but do it in a
> separate patch.

Ok, Let's postpone custom version.
-- 
Isaku Yamahata <isaku.yamahata@...el.com>