patches for DPDK stable branches
 help / color / mirror / Atom feed
From: Kevin Traynor <ktraynor@redhat.com>
To: Fabio Pricoco <fabio.pricoco@intel.com>,
	"Richardson, Bruce" <bruce.richardson@intel.com>
Cc: Jacob Keller <jacob.e.keller@intel.com>,
	Soumyadeep Hore <soumyadeep.hore@intel.com>,
	dpdk stable <stable@dpdk.org>
Subject: Re: patch 'net/ice/base: fix iteration of TLVs in Preserved Fields Area' has been queued to stable release 21.11.9
Date: Tue, 3 Dec 2024 14:12:32 +0000	[thread overview]
Message-ID: <a56fbe04-3b2f-40bd-a0ee-23505c9fd5b9@redhat.com> (raw)
In-Reply-To: <20241127171916.690404-30-ktraynor@redhat.com>

On 27/11/2024 18:17, Kevin Traynor wrote:
> Hi,
> 
> FYI, your patch has been queued to stable release 21.11.9
> 

Hi Bruce/Fabio,

This patch is causing a build issue for CentOS7 as
__builtin_add_overflow is not available on GCC 4.8.5.

Though CentOS7 has just gone EoL, I don't want to break the build with
the last 21.11 LTS release.

I think we could replace the check_add_overflow/builtin with something
like below (untested), but if the patch isn't important for 21.11
branch, then it's probably safer to drop it.

How much is the patch needed for 21.11 branch ?

thanks,
Kevin.

--- a/drivers/net/ice/base/ice_nvm.c
+++ b/drivers/net/ice/base/ice_nvm.c
@@ -427,6 +427,4 @@ enum ice_status ice_read_sr_word(struct ice_hw *hw,
u16 offset, u16 *data)
 }

-#define check_add_overflow __builtin_add_overflow
-
 /**
  * ice_get_pfa_module_tlv - Reads sub module TLV from NVM PFA
@@ -458,5 +456,6 @@ ice_get_pfa_module_tlv(struct ice_hw *hw, u16
*module_tlv, u16 *module_tlv_len,
        }

-       if (check_add_overflow(pfa_ptr, (u16)(pfa_len - 1), &max_tlv)) {
+       max_tlv = pfa_ptr + pfa_len - 1;
+       if (pfa_ptr > max_tlv) {
                ice_debug(hw, ICE_DBG_INIT, "PFA starts at offset %u.
PFA length of %u caused 16-bit arithmetic overflow.\n",
                                  pfa_ptr, pfa_len);
@@ -475,4 +474,5 @@ ice_get_pfa_module_tlv(struct ice_hw *hw, u16
*module_tlv, u16 *module_tlv_len,
        while (next_tlv < max_tlv) {
                u16 tlv_sub_module_type;
+               u16 curr_tlv;
                u16 tlv_len;

@@ -499,6 +499,7 @@ ice_get_pfa_module_tlv(struct ice_hw *hw, u16
*module_tlv, u16 *module_tlv_len,
                }

-               if (check_add_overflow(next_tlv, (u16)2, &next_tlv) ||
-                   check_add_overflow(next_tlv, tlv_len, &next_tlv)) {
+               curr_tlv = next_tlv;
+               next_tlv = next_tlv + tlv_len + 2;
+               if (curr_tlv > next_tlv) {
                        ice_debug(hw, ICE_DBG_INIT, "TLV of type %u and
length 0x%04x caused 16-bit arithmetic overflow. The PFA starts at
0x%04x and has length of 0x%04x\n",
                                          tlv_sub_module_type, tlv_len,
pfa_ptr, pfa_len);


> Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
> It will be pushed if I get no objections before 12/02/24. So please
> shout if anyone has objections.
> 
> Also note that after the patch there's a diff of the upstream commit vs the
> patch applied to the branch. This will indicate if there was any rebasing
> needed to apply to the stable branch. If there were code changes for rebasing
> (ie: not only metadata diffs), please double check that the rebase was
> correctly done.
> 
> Queued patches are on a temporary branch at:
> https://github.com/kevintraynor/dpdk-stable
> 
> This queued commit can be viewed at:
> https://github.com/kevintraynor/dpdk-stable/commit/ebbecb19ccdb0df88227e69b2a83ae2ee79a2d19
> 
> Thanks.
> 
> Kevin
> 
> ---
> From ebbecb19ccdb0df88227e69b2a83ae2ee79a2d19 Mon Sep 17 00:00:00 2001
> From: Fabio Pricoco <fabio.pricoco@intel.com>
> Date: Fri, 23 Aug 2024 09:56:42 +0000
> Subject: [PATCH] net/ice/base: fix iteration of TLVs in Preserved Fields Area
> 
> [ upstream commit dcb760bf0f951b404bce33a1dd14906154b58c75 ]
> 
> The ice_get_pfa_module_tlv() function iterates over the Preserved Fields
> Area to read data from the Shadow RAM, including the Part Board Assembly
> data, among others.
> 
> If the specific TLV being requested is not found in the current NVM, the
> code will read past the end of the PFA, misinterpreting the last word of
> the PFA and the word just after the PFA as another TLV. This typically
> results in one extra iteration before the length check of the while loop
> is triggered.
> 
> Correct the logic for determining the maximum PFA offset to include the
> extra last word. Additionally, make the driver robust against overflows
> by using check_add_overflow. This ensures that even if the NVM provides
> bogus data, the driver will not overflow, and will instead log a useful
> warning message. The check for whether the TLV length exceeds the PFA
> length is also removed, in favor of relying on the overflow warning
> instead.
> 
> Fixes: 5d0b7b5fc491 ("net/ice/base: add read PBA module function")
> 
> Signed-off-by: Fabio Pricoco <fabio.pricoco@intel.com>
> Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
> Signed-off-by: Soumyadeep Hore <soumyadeep.hore@intel.com>
> Acked-by: Bruce Richardson <bruce.richardson@intel.com>
> ---
>  drivers/net/ice/base/ice_nvm.c | 36 ++++++++++++++++++++++------------
>  1 file changed, 24 insertions(+), 12 deletions(-)
> 
> diff --git a/drivers/net/ice/base/ice_nvm.c b/drivers/net/ice/base/ice_nvm.c
> index 48e0d418e2..c5a3eddebf 100644
> --- a/drivers/net/ice/base/ice_nvm.c
> +++ b/drivers/net/ice/base/ice_nvm.c
> @@ -427,4 +427,6 @@ enum ice_status ice_read_sr_word(struct ice_hw *hw, u16 offset, u16 *data)
>  }
>  
> +#define check_add_overflow __builtin_add_overflow
> +
>  /**
>   * ice_get_pfa_module_tlv - Reads sub module TLV from NVM PFA
> @@ -443,6 +445,5 @@ ice_get_pfa_module_tlv(struct ice_hw *hw, u16 *module_tlv, u16 *module_tlv_len,
>  {
>  	enum ice_status status;
> -	u16 pfa_len, pfa_ptr;
> -	u32 next_tlv;
> +	u16 pfa_len, pfa_ptr, next_tlv, max_tlv;
>  
>  	status = ice_read_sr_word(hw, ICE_SR_PFA_PTR, &pfa_ptr);
> @@ -456,9 +457,21 @@ ice_get_pfa_module_tlv(struct ice_hw *hw, u16 *module_tlv, u16 *module_tlv_len,
>  		return status;
>  	}
> -	/* Starting with first TLV after PFA length, iterate through the list
> +
> +	if (check_add_overflow(pfa_ptr, (u16)(pfa_len - 1), &max_tlv)) {
> +		ice_debug(hw, ICE_DBG_INIT, "PFA starts at offset %u. PFA length of %u caused 16-bit arithmetic overflow.\n",
> +				  pfa_ptr, pfa_len);
> +		return ICE_ERR_INVAL_SIZE;
> +	}
> +
> +	/* The Preserved Fields Area contains a sequence of TLVs which define
> +	 * its contents. The PFA length includes all of the TLVs, plus its
> +	 * initial length word itself, *and* one final word at the end of all
> +	 * of the TLVs.
> +	 *
> +	 * Starting with first TLV after PFA length, iterate through the list
>  	 * of TLVs to find the requested one.
>  	 */
>  	next_tlv = pfa_ptr + 1;
> -	while (next_tlv < ((u32)pfa_ptr + pfa_len)) {
> +	while (next_tlv < max_tlv) {
>  		u16 tlv_sub_module_type;
>  		u16 tlv_len;
> @@ -477,8 +490,4 @@ ice_get_pfa_module_tlv(struct ice_hw *hw, u16 *module_tlv, u16 *module_tlv_len,
>  			break;
>  		}
> -		if (tlv_len > pfa_len) {
> -			ice_debug(hw, ICE_DBG_INIT, "Invalid TLV length.\n");
> -			return ICE_ERR_INVAL_SIZE;
> -		}
>  		if (tlv_sub_module_type == module_type) {
>  			if (tlv_len) {
> @@ -489,8 +498,11 @@ ice_get_pfa_module_tlv(struct ice_hw *hw, u16 *module_tlv, u16 *module_tlv_len,
>  			return ICE_ERR_INVAL_SIZE;
>  		}
> -		/* Check next TLV, i.e. current TLV pointer + length + 2 words
> -		 * (for current TLV's type and length)
> -		 */
> -		next_tlv = next_tlv + tlv_len + 2;
> +
> +		if (check_add_overflow(next_tlv, (u16)2, &next_tlv) ||
> +		    check_add_overflow(next_tlv, tlv_len, &next_tlv)) {
> +			ice_debug(hw, ICE_DBG_INIT, "TLV of type %u and length 0x%04x caused 16-bit arithmetic overflow. The PFA starts at 0x%04x and has length of 0x%04x\n",
> +					  tlv_sub_module_type, tlv_len, pfa_ptr, pfa_len);
> +			return ICE_ERR_INVAL_SIZE;
> +		}
>  	}
>  	/* Module does not exist */


  reply	other threads:[~2024-12-03 14:12 UTC|newest]

Thread overview: 130+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-11-27 17:17 patch 'net/ionic: fix build with Fedora Rawhide' " Kevin Traynor
2024-11-27 17:17 ` patch 'eal/x86: fix 32-bit write combining store' " Kevin Traynor
2024-11-27 17:17 ` patch 'examples/eventdev: fix queue crash with generic pipeline' " Kevin Traynor
2024-11-27 17:17 ` patch 'crypto/dpaa2_sec: fix memory leak' " Kevin Traynor
2024-11-27 17:17 ` patch 'common/dpaax/caamflib: fix PDCP SNOW-ZUC watchdog' " Kevin Traynor
2024-11-27 17:17 ` patch 'dev: fix callback lookup when unregistering device' " Kevin Traynor
2024-11-27 17:17 ` patch 'bpf: fix free function mismatch if convert fails' " Kevin Traynor
2024-11-27 17:17 ` patch 'baseband/la12xx: fix use after free in modem config' " Kevin Traynor
2024-11-27 17:17 ` patch 'crypto/bcmfs: fix free function mismatch' " Kevin Traynor
2024-11-27 17:17 ` patch 'dma/idxd: fix free function mismatch in device probe' " Kevin Traynor
2024-11-27 17:17 ` patch 'event/cnxk: fix free function mismatch in port config' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/e1000: fix use after free in filter flush' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/sfc: fix use after free in debug logs' " Kevin Traynor
2024-11-27 17:17 ` patch 'raw/ifpga/base: fix use after free' " Kevin Traynor
2024-11-27 17:17 ` patch 'raw/ifpga: fix free function mismatch in interrupt config' " Kevin Traynor
2024-11-27 17:17 ` patch 'examples/vhost: fix free function mismatch' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/nfb: fix use after free' " Kevin Traynor
2024-11-27 17:17 ` patch 'power: enable CPPC' " Kevin Traynor
2024-11-27 17:17 ` patch 'fib6: add runtime checks in AVX512 lookup' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/cnxk: fix Rx timestamp handling for VF' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/cnxk: fix Rx offloads to handle timestamp' " Kevin Traynor
2024-11-27 17:17 ` patch 'common/cnxk: fix base log level' " Kevin Traynor
2024-11-27 17:17 ` patch 'vhost: fix offset while mapping log base address' " Kevin Traynor
2024-11-27 17:17 ` patch 'vdpa: update used flags in used ring relay' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/virtio-user: reset used index counter' " Kevin Traynor
2024-11-27 17:17 ` patch 'fib: fix AVX512 lookup' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/e1000: fix link status crash in secondary process' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/iavf: fix crash when link is unstable' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/ice/base: fix link speed for 200G' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/ice/base: fix iteration of TLVs in Preserved Fields Area' " Kevin Traynor
2024-12-03 14:12   ` Kevin Traynor [this message]
2024-12-03 15:10     ` Richardson, Bruce
2024-12-03 15:36       ` Kevin Traynor
2024-12-03 15:39         ` Richardson, Bruce
2024-12-03 15:43           ` Kevin Traynor
2024-11-27 17:17 ` patch 'net/ixgbe/base: fix unchecked return value' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/i40e/base: fix setting flags in init function' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/i40e/base: fix misleading debug logs and comments' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/i40e/base: fix blinking X722 with X557 PHY' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/i40e/base: fix DDP loading with reserved track ID' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/i40e/base: fix repeated register dumps' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/i40e/base: fix unchecked return value' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/i40e/base: fix loop bounds' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/i40e: fix AVX-512 pointer copy on 32-bit' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/ice: " Kevin Traynor
2024-11-27 17:17 ` patch 'net/iavf: " Kevin Traynor
2024-11-27 17:17 ` patch 'net/tap: avoid memcpy with null argument' " Kevin Traynor
2024-11-27 17:17 ` patch 'app/testpmd: remove unnecessary cast' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/pcap: set live interface as non-blocking' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/ena: revert redefining memcpy' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/hns3: remove some basic address dump' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/hns3: fix dump counter of registers' " Kevin Traynor
2024-11-27 17:17 ` patch 'ethdev: fix overflow in descriptor count' " Kevin Traynor
2024-11-27 17:17 ` patch 'bus/dpaa: fix PFDRs leaks due to FQRNIs' " Kevin Traynor
2024-11-27 17:17 ` patch 'net/dpaa: fix typecasting channel ID' " Kevin Traynor
2024-11-27 17:17 ` patch 'bus/dpaa: fix VSP for 1G fm1-mac9 and 10' " Kevin Traynor
2024-11-27 17:17 ` patch 'bus/dpaa: fix the fman details status' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/dpaa: fix reallocate mbuf handling' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/memif: fix buffer overflow in zero copy Rx' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/tap: restrict maximum number of MP FDs' " Kevin Traynor
2024-11-27 17:18 ` patch 'ethdev: verify queue ID in Tx done cleanup' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/hns3: verify reset type from firmware' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/pcap: fix blocking Rx' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/ice/base: add bounds check' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/ice/base: fix VLAN replay after reset' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/iavf: preserve MAC address with i40e PF Linux driver' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/mlx5: workaround list management of Rx queue control' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/mlx5: fix number of supported flex parsers' " Kevin Traynor
2024-11-27 17:18 ` patch 'app/testpmd: remove flex item init command leftover' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/mlx5: fix next protocol validation after flex item' " Kevin Traynor
2024-11-27 17:18 ` patch 'build: remove version check on compiler links function' " Kevin Traynor
2024-11-27 17:18 ` patch 'hash: fix thash LFSR initialization' " Kevin Traynor
2024-11-27 17:18 ` patch 'dmadev: fix potential null pointer access' " Kevin Traynor
2024-11-27 17:18 ` patch 'power: fix mapped lcore ID' " Kevin Traynor
2024-12-09  7:00   ` David Marchand
2024-12-09  9:21     ` Tummala, Sivaprasad
2024-12-09 10:25       ` Kevin Traynor
2024-11-27 17:18 ` patch 'net/netvsc: fix using Tx queue higher than Rx queues' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/hns3: restrict tunnel flow rule to one header' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/ice: detect stopping a flow director queue twice' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/ixgbe: fix link status delay on FreeBSD' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/mvneta: fix possible out-of-bounds write' " Kevin Traynor
2024-11-27 17:18 ` patch 'crypto/openssl: fix 3DES-CTR with big endian CPUs' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/mlx5: fix memory leak in metering' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/mlx5: fix reported Rx/Tx descriptor limits' " Kevin Traynor
2024-11-27 17:18 ` patch 'app/dumpcap: remove unused struct array' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/nfp: fix link change return value' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/mlx5: fix non full word sample fields in flex item' " Kevin Traynor
2024-11-27 17:18 ` patch 'bus/fslmc: fix Coverity warnings in QBMAN' " Kevin Traynor
2024-11-27 17:18 ` patch 'test/bonding: remove redundant info query' " Kevin Traynor
2024-11-27 17:18 ` patch 'examples/ntb: check info query return' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/netvsc: force Tx VLAN offload on 801.2Q packet' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/vmxnet3: fix crash after configuration failure' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/txgbe: fix SWFW mbox' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/txgbe: fix VF-PF mbox interrupt' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/txgbe: remove outer UDP checksum capability' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/txgbe: fix driver load bit to inform firmware' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/ngbe: " Kevin Traynor
2024-11-27 17:18 ` patch 'net/ngbe: reconfigure more MAC Rx registers' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/ngbe: fix interrupt lost in legacy or MSI mode' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/ngbe: restrict configuration of VLAN strip offload' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/hns3: fix error code for repeatedly create counter' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/hns3: fix fully use hardware flow director table' " Kevin Traynor
2024-11-27 17:18 ` patch 'event/octeontx: fix possible integer overflow' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/bnxt/tf_core: fix Thor TF EM key size check' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/bnxt: fix reading SFF-8436 SFP EEPROMs' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/bnxt: fix TCP and UDP checksum flags' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/bnxt: fix bad action offset in Tx BD' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/bnx2x: remove dead conditional' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/bnx2x: fix always true expression' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/bnx2x: fix possible infinite loop at startup' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/bnx2x: fix duplicate branch' " Kevin Traynor
2024-11-27 17:18 ` patch 'common/cnxk: fix build on Ubuntu 24.04' " Kevin Traynor
2024-11-27 17:18 ` patch 'examples/l2fwd-event: fix spinlock handling' " Kevin Traynor
2024-11-27 17:18 ` patch 'eventdev: fix possible array underflow/overflow' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/mlx5: fix shared queue port number in vector Rx' " Kevin Traynor
2024-11-27 17:18 ` patch 'common/mlx5: fix misalignment' " Kevin Traynor
2024-11-27 17:18 ` patch 'bus/dpaa: fix lock condition during error handling' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/iavf: add segment-length check to Tx prep' " Kevin Traynor
2024-11-27 17:18 ` patch 'net/i40e: check register read for outer VLAN' " Kevin Traynor
2024-11-27 17:19 ` patch 'app/procinfo: fix leak on exit' " Kevin Traynor
2024-11-27 17:19 ` patch 'member: fix choice of bucket for displacement' " Kevin Traynor
2024-11-27 17:19 ` patch 'app/testpmd: fix aged flow destroy' " Kevin Traynor
2024-11-27 17:19 ` patch 'test/bonding: fix loop on members' " Kevin Traynor
2024-11-27 17:19 ` patch 'test/bonding: fix MAC address comparison' " Kevin Traynor
2024-11-27 17:19 ` patch 'test/event: avoid duplicate initialization' " Kevin Traynor
2024-11-27 17:19 ` patch 'test/eal: fix loop coverage for alignment macros' " Kevin Traynor
2024-11-27 17:19 ` patch 'test/eal: fix lcore check' " Kevin Traynor
2024-11-27 17:19 ` patch 'app/testpmd: remove redundant policy action condition' " Kevin Traynor
2024-11-27 17:19 ` patch 'doc: correct definition of stats per queue feature' " Kevin Traynor

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=a56fbe04-3b2f-40bd-a0ee-23505c9fd5b9@redhat.com \
    --to=ktraynor@redhat.com \
    --cc=bruce.richardson@intel.com \
    --cc=fabio.pricoco@intel.com \
    --cc=jacob.e.keller@intel.com \
    --cc=soumyadeep.hore@intel.com \
    --cc=stable@dpdk.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).